When the ABB FIA Formula E World Championship launched its first race through Beijing’s Olympic Park in 2014, the idea of all-electric motorsport still bordered on experimental. Batteries couldn’t yet last a full race, and drivers had to switch cars mid-competition. Just over a decade later, Formula E has evolved into a global entertainment brand broadcast in 150 countries, driving both technological innovation and cultural change in sport.  

“Gen4, that’s to come next year,” says Dan Cherowbrier, Formula E’s chief technology and information officer. “You will see a really quite impressive car that starts us to question whether EV is there. It’s actually faster—it’s actually more than traditional [internal combustion engines] ICE.” 

That acceleration isn’t just happening on the track. Formula E’s digital transformation, powered by its partnership with Infosys, is redefining what it means to be a fan. “It’s a movement to make motor sport accessible and exciting for the new generation,” says principal technologist at Infosys, Rohit Agnihotri. 

From real-time leaderboards and predictive tools to personalized storylines that adapt to what individual fans care most about—whether it’s a driver rivalry or battery performance—Formula E and Infosys are using AI-powered platforms to create fan experiences as dynamic as the races themselves. “Technology is not just about meeting expectations; it’s elevating the entire fan experience and making the sport more inclusive,” says Agnihotri.  

AI is also transforming how the organization itself operates. “Historically, we would be going around the company, banging on everyone’s doors and dragging them towards technology, making them use systems, making them move things to the cloud,” Cherowbrier notes. “What AI has done is it’s turned that around on its head, and we now have people turning up, banging on our door because they want to use this tool, they want to use that tool.” 

As audiences diversify and expectations evolve, Formula E is also a case study in sustainable innovation. Machine learning tools now help determine the most carbon-optimal way to ship batteries across continents, while remote broadcast production has sharply reduced travel emissions and democratized the company’s workforce. These advances show how digital intelligence can expand reach without deepening carbon footprints. 

For Cherowbrier, this convergence of sport, sustainability, and technology is just the beginning. With its data-driven approach to performance, experience, and impact, Formula E is offering a glimpse into how entertainment, innovation, and environmental responsibility can move forward in tandem. 

“Our goal is clear,” says Agnihotri. “Help Formula E be the most digital and sustainable motor sport in the world. The future is electric, and with AI, it’s more engaging than ever.” 

This episode of Business Lab is produced in partnership with Infosys. 

Full Transcript:  

Megan Tatum: From MIT Technology Review, I’m Megan Tatum, and this is Business Lab, the show that helps business leaders make sense of new technologies coming out of the lab, and into the marketplace.  

The ABB FIA Formula E World Championship, the world’s first all-electric racing series, made its debut in the grounds of the Olympic Park in Beijing in 2014. A little more than 10 years later, it’s a global entertainment brand with 10 teams, 20 drivers, and broadcasts in 150 countries. Technology is central to how Formula E is navigating that scale and to how it’s delivering more powerful personalized experiences.  

Two words for you: elevated fandom.  

My guests today are Rohit Agnihotri, principal technologist at Infosys, and Dan Cherowbrier, CTIO of Formula E.  

This episode is produced in partnership with Infosys.  

Welcome, Rohit and Dan. 

Dan Cherowbrier: Hi. Thanks for having us. 

Megan: Dan, as I mentioned there, the first season of the ABB FIA Formula E World Championship launched in 2014. Can you talk us through how the first all-electric motor sport has evolved in the last decade? How has it changed in terms of its scale, the markets it operates in, and also, its audiences, of course? 

Dan: When Formula E launched back in 2014, there were hardly any domestic EVs on the road. And probably if you’re from London, the ones you remember are the hybrid Priuses; that was what we knew of really. And at the time, they were unable to get a battery big enough for a car to do a full race. So the first generation of car, the first couple of seasons, the driver had to do a pit stop midway through the race, get out of one car, and get in another car, and then carry on, which sounds almost farcical now, but it’s what you had to do then to drive innovation, is to do that in order to go to the next stage. 

Then in Gen2, that came up four years later, they had a battery big enough to start full races and start to actually make it a really good sport. Gen3, they’re going for some real speeds and making it happen. Gen4, that’s to come next year, you’ll see acceleration in line with Formula One. I’ve been fortunate enough to see some of the testing. You will see a really quite impressive car that starts us to question whether EV is there. It’s actually faster, it’s actually more than traditional ICE. 

That’s the tech of the car. But then, if you also look at the sport and how people have come to it and the fans and the demographic of the fans, a lot has changed in the last 11 years. We were out to enter season 12. In the last 11 years, we’ve had a complete democratization of how people access content and what people want from content. And as a new generation of fan coming through. This new generation of fan is younger. They’re more gender diverse. We have much closer to 50-50 representation in our fan base. And they want things personalized, and they’re very demanding about how they want it and the experience they expect. No longer are you just able to give them one race and everybody watches the same thing. We need to make things for them. You see that sort of change that’s come through in the last 11 years. 

Megan: It’s a huge amount of change in just over a decade, isn’t it? To navigate. And I wonder, Rohit, what was the strategic plan for Infosys when associating with Formula E? What did Infosys see in partnering with such a young sport? 

Rohit: Yeah. That’s a great question, Megan. When we looked at Formula E, we didn’t just see a racing championship. We saw the future. A sport, that’s electric, sustainable, and digital first. That’s exactly where Infosys wants to be, at the intersection of technology, innovation, and purpose. Our plan has three big goals. First, grow the fan base. Formula E wants to reach 500 million fans by 2030. That is not just a number. It’s a movement to make motor sport accessible and exciting for the new generation. To make that happen, we are building an AI-powered platform that gives personalized content to the fans, so that every fan feels connected and valued. Imagine a fan in Tokyo getting race insights tailored for their favorite driver, while another in London gets a sustainability story that matters to him. That’s the level of personalization we are aiming for. 

Second, bringing technology innovation. We have already launched the Stats Centre, which turns race data into interactive stories. And soon, Race Centre will take this to the next level with real time leaderboards to the race or tracks, overtakes, attack mode timelines, and even AI generated live commentary. Fans will not just watch, they will interact, predict podium finishes, and share their views globally. And third, supports sustainability. Formula E is already net-zero, but now their goal is to cut carbon by 45% by 2030. We’ll be enabling that through AI-driven sustainability, data management, tracking every watt of energy, every logistics decision. and modeling scenarios to make racing even greener. Partnering with a young sport gives us a chance to shape its digital future and show how technology can make racing exciting and responsible. For us, Formula E is not just a sport, it’s a statement about where the world is headed. 

Megan: Fantastic. 500 million fans, that’s a huge number, isn’t it? And with more scale often comes a kind of greater expectation. Dan, I know you touched on this a little in your first question, but what is it that your fans now really want from their interactions? Can you talk a bit more about what experiences they’re looking for? And also, how complex that really is to deliver that as well? 

Dan: I think a really telling thing about the modern day fan is I probably can’t tell you what they want from their experiences, because it’s individual and it’s unique for each of them. 

Megan: Of course. 

Dan: And it’s changing and it’s changing so fast. What somebody wants this month is going to be different from what they want in a couple of months’ time. And we’re having to learn to adapt to that. My CTO title, we often put focus on the technology in the middle of it. That’s what the T is. Actually, if you think about it, it’s continual transformation officer. You are constantly trying to change what you deliver and how you deliver it. Because if fans come through, they find new experiences, they find that in other sports. Sometimes not in sports, they find it outside, and then they’re coming in, and they expect that from you. So how can we make them more part of the sport, more personalized experience, get to know the athletes and the personalities and the characters within it? We’re a very technology centric sport. A lot of motor sport is, but really, people want to see people, right? And even when it’s technology, they want to see people interacting with technology, and it’s how do you get that out to show people. 

Megan: Yeah, it’s no mean feat. Rohit, you’ve worked with brands on delivering these sort of fan experiences across different sports. Is motor sports perhaps more complicated than others, given that fans watch racing for different reasons than just a win? They could be focused on team dynamics, a particular driver, the way the engine is built, and so on and so forth. How does motor sports compare and how important is it therefore, that Formula E has embraced technology to manage expectations? 

Rohit: Yeah, that’s an interesting point. Motor sports are definitely more complex than other sports. Fans don’t just care about who wins, they care about how some follow team strategies, others love driver rivalries, and many are fascinated by the car technology. Formula E adds another layer, sustainability and electric innovation. This makes personalization really important. Fans want more than results. They want stories and insights. Formula E understood this early and embraced technology. 

Think about the data behind a single race, lap times, energy usage, battery performance, attack mode activation, pit strategies, it’s a lot of data. If you just show the raw numbers, it’s overwhelming. But with Infosys Topaz, we turn that into simple and engaging stories. Fans can see how a driver fought back from 10th place to finish on the podium, or how a team managed energy better to gain an edge. And for new fans, we are adding explainer videos and interactive tools in the Race Center, so that they can learn about their sport easily. This is important because Formula E is still young, and many fans are discovering it for the first time. Technology is not just about meeting expectations; it’s elevating the entire fan experience and making the sport more inclusive. 

Megan: There’s an awful lot going on there. What are some of the other ways that Formula E has already put generative AI and other emerging technologies to use? Dan, when we’ve spoken about the demand for more personalized experiences, for example. 

Dan: I see the implementation of AI for us in three areas. We have AI within the sport. That’s in our DNA of the sport. Now, each team is using that, but how can we use that as a championship as well? How do we make it a competitive landscape? Now, we have AI that is in the fan-facing product. That’s what we’re working heavily on Infosys with, but we also have it in our broadcast product. As an example, you might have heard of a super slow-mo camera. A super slow-mo camera is basically, by taking three cameras and having them in exactly the same place so that you get three times the frame rate, and then you can do a slow-motion shot from that. And they used to be really expensive. Quite bulky cameras to put in. We are now using AI to take a traditional camera and interpolate between two frames to make it into a super slow image, and you wouldn’t really know the difference. Now, the joy of that, it means every camera can now be a super slow-mo camera. 

Megan: Wow. 

Dan: In other ways, we use it a little bit in our graphics products, and we iterate and we use it for things like showing driver audio. When the driver is speaking to his engineer or her engineer in the garage, we show that text now on screen. We do that using AI. We use AI to pick out the difference between the driver and another driver and the team engineer or the team principal and show that in a really good way. 

And we wouldn’t be able to do that. We’re not big enough to have a team of 24 people on stenographers typing. We have to use AI to be able to do that. That’s what’s really helped us grow. And then the last one is, how we use it in our business. Because ultimately, as we’ve got the fans, we’ve got the sport, but we also are running a business and we have to pick up these racetracks and move them around the world, and we have all these staff who have to get places. We have insurance who has to do all that kind of stuff, and we use it heavily in that area, particularly when it comes to what has a carbon impact for us. 

So things like our freight and our travel. And we are using the AI tools to tell us, a battery for instance, should we fly it? Should we send it by sea freight? Should we send it by row freight? Or should we just have lots of them? And that sort of depends. Now, a battery, if it was heavy, you’d think you probably wouldn’t fly it. But actually, because of the materials in it, because of the source materials that make it, we’re better off flying it. We’ve used AI to work through all those different machinations of things that would be too difficult to do at speed for a person. 

Megan: Well, sounds like there’s some fascinating things going on. I mean, of course, for a global brand, there is also the challenge of working in different markets. You mentioned moving everything around the world there. Each market with its own legal frameworks around data privacy, AI. How has technology also helped you navigate all of that, Dan? 

Dan: The other really interesting thing about AI is… I’ve worked in technology leadership roles for some time now. And historically, we would be going around the company, banging on everyone’s doors and dragging them towards technology, making them use systems, making them move things to the cloud and things like that. What AI has done is it’s turned that around on its head, and we now have people turning up, banging on our door because they want to use this tool, they want to use that tool. And we’re trying to accommodate all of that and it’s a great pleasure to see people that are so keen. AI is driving the tech adoption in general, which really helps the business. 

Megan: Dan, as the world’s first all-electric motor sport series, sustainability is obviously a real cornerstone of what Formula E is looking to do. Can you share with us how technology is helping you to achieve some of your ambitions when it comes to sustainability? 

Dan: We’ve been the only sport with a certified net-zero pathway, and we have to stay that part. It’s a really core fundamental part of our DNA. I sit on our management team here. There is a sustainability VP that sits there as well, who checks and challenges everything we do. She looks at the data centers we use, why we use them, why we’ve made the decisions we’ve made, to make sure that we’re making them all for the right reasons and the right ways. We specifically embed technology in a couple of ways. One is, we mentioned a little bit earlier, on our freight. Formula E’s freight for the whole championship is probably akin to one Formula One team, but it’s still by far, our biggest contributor to our impact. So we look about how we can make sure that we’ve refined that to get the minimum amount of air freight and sea freight, and use local wherever we can. That’s also part of our pledge about investing in the communities that we race in. 

The second then is about our staff travel. And we’ve done a really big piece of work over the last four to five years, partly accelerated through the covid-19 era actually, of doing remote working and remote TV production. Used to be traditionally, you would fly a hundred plus people out to racetracks, and then they would make the television all on site in trucks, and then they would be satellite distributed out of the venue. Now, what we do is we put in some internet connections, dual and diverse internet connections, and we stream every single camera back. 

Megan: Right. 

Dan: That means on site, we only need camera operators. Some of them actually, are remotely operated anyway, but we need camera operators, and then some engineering teams to just keep everything running. And then back in our home base, which is in London, in the UK, we have our remote production center where we layer on direction, graphics, audio, replay, team radio, all of those bits that break the color and make the program and add to that significant body of people. We do that all remotely now. Really interesting actually, a bit. So that’s the carbon sustainability story, but there is a further ESG piece that comes out of it and we haven’t really accommodated when we went into it, is the diversity in our workforce by doing that. We were discovering that we had quite a young, equally diverse workforce until around the age of 30. And then once that happened, then we were finding we were losing women, and that’s really because they didn’t want to travel. 

Megan: Right. 

Dan: And that’s the age of people starting to have children, and things were starting to change. And then we had some men that were traveling instead, and they weren’t seeing their children and it was sort of dividing it unnecessarily. But by going remote, by having so much of our people able to remotely… Or even if they do have to travel, they’re not traveling every single week. They’re now doing that one in three. They’re able to maintain the careers and the jobs they want to do, whilst having a family lifestyle. And it also just makes a better product by having people in that environment. 

Megan: That’s such an interesting perspective, isn’t it? It’s a way of environmental sustainability intersects with social sustainability. And Rohit, and your work are so interesting. And Rohit, can you share any of the ways that Infosys has worked with Formula E, in terms of the role of technology as we say, in furthering those ambitions around sustainability? 

Rohit: Yeah. Infosys understands that sustainability is at the heart of Formula E, and it’s a big part of why this partnership matters. Formula E is already net-zero certified, but now, they have an ambitious goal to cut carbon emissions by 45%. Infosys is helping in two ways. First, we have built AI-powered sustainability data tools that make carbon reporting accurate and traceable. Every watt of energy, every logistic decision, every material use can be tracked. Second, we use predictive analytics to model scenarios, like how changing race logistics or battery technology impact emissions so Formula E can make smarter, greener decisions. For us, it’s about turning sustainability from a report into an action plan, and making Formula E a global leader in green motor sport. 

Megan: And in April 2025, Formula E working with Infosys launched its Stats Centre, which provides fans with interactive access to the performances of their drivers and teams, key milestones and narratives. I know you touched on this before, but I wonder if you could tell us a bit more about the design of that platform, Rohit, and how it fits into Formula E’s wider plans to personalize that fan experience? 

Rohit: Sure. The Stats Centre was a big step forward. Before this, fans had access to basic statistics on the website and the mobile app, but nothing told the full story and we wanted to change that. Built on Infosys Topaz, the Stats Centre uses AI to turn race data into interactive stories. Fans can explore key stat cards that adapt to race timelines, and even chat with an AI companion to get instant answers. It’s like having a person race analyst at your fingertips. And we are going further. Next year, we’ll launch Race Centre. It’ll have live data boards, 2D track maps showing every driver’s position, overtakes and more attack timelines, and AI-generated commentary. Fans can predict podium finishes, vote for the driver of the race, and share their views on social media. Plus, we are adding video explainers for new fans, covering rules, strategies, and car technology. Our goal is simple: make every moment exciting and easy to understand. Whether you are a hardcore fan or someone watching Formula E for the first time, you’ll feel connected and informed. 

Megan: Fantastic. Sounds brilliant. And as you’ve explained, Dan, leveraging data and AI can come with these huge benefits when it comes to the depth of fan experience that you can deliver, but it can also expose you to some challenges. How are you navigating those at Formula E? 

Dan: The AI generation has presented two significant challenges to us. One is that traditional SEO, traditional search engine optimization, goes out the window. Right? You are now looking at how do we design and build our systems and how do we populate them with the right content and the right data, so that the engines are picking it up correctly and displaying it? The way that the foundational models are built and the speed and the cadence of which they’re updated, means quite often… We’re a very fast-changing organization. We’re a fast-changing product. Often, the models don’t keep up. And that’s because they are a point in time when they were trained. And that’s something that the big organizations, the big tech organizations will fix with time. But for now, what we have to do is we have to learn about how we can present our fan-facing, web-facing products to show that correctly. That’s all about having really accurate first-party content, effectively earned media. That’s the piece we need to do. 

Then the second sort of challenge is sadly, whilst these tools are available to all of us, and we are using them effectively, so are another part of the technology landscape, and that is the cybersecurity basically they come with. If you look at the speed of the cadence and severity of hacks that are happening now, it’s just growing and growing and growing, and that’s because they have access to these tools too. And we’re having to really up our game and professionalize. And that’s really hard for an innovative organization. You don’t want to shut everything down. You don’t want to protect everything too much because you want people to be able to try new things. Right? If I block everything to only things that the IT team had heard of, we’d never get anything new in, and it’s about getting that balance right. 

Megan: Right. 

Dan: Rohit, you probably have similar experiences? 

Megan: How has Infosys worked with Formula E to help it navigate some of that, Rohit? 

Rohit: Yeah. Infosys has helped Formula E tackle some of the challenges in three key ways, simplify complex race data into engaging fan experience through platforms like Stats Centre, building a secure and scalable cloud data backbone for the real-time insights, and enabling sustainability goals with AI-driven carbon tracking and predictive analytics. This solution makes the sport interactive, more digital, and more responsible. 

Megan: Fantastic. I wondered if we could close with a bit of a future forward look. Can you share with us any innovations on the horizon at Formula E that you are really excited about, Dan? 

Dan: We have mentioned the Race Centre is going to launch in the next couple of months, but the really exciting thing for me is we’ve got an amazing season ahead of us. It’s the last season of our Gen3 car, with 10 really exciting teams on the grid. We are going at speed with our tech innovation roadmap and what our fans want. And we’re building up towards our Gen4 car, which will come out for season 13 in a year’s time. That will get launched in 2026, and I think it will be a game changer in how people perceive electric motor sport and electric cars in general. 

Megan: It sounds like there’s all sorts of exciting things going on. And Rohit too, what’s coming up via this partnership that you are really looking forward to sharing with everyone? 

Rohit: Two things stand out for me. First is the AI-powered fan data platform that I’ve already spoken about. Second is the launch of Race Centre. It’s going to change how fans experience live racing. And beyond final engagement, we are helping Formula E lead in sustainability with AI tools that model carbon impact and optimize logistics. This means every race can be smarter and greener. Our goal is clear: help Formula E be the most digital and sustainable motor sport in the world. The future is electric, and with AI, it’s more engaging than ever. 

Megan: Fantastic. Thank you so much, both. That was Rohit Agnihotri, principal technologist at Infosys, and Dan Cherowbrier, CITO of Formula E, whom I spoke with from Brighton, England.  

That’s it for this episode of Business Lab. I’m your host, Megan Tatum. I’m a contributing editor and host for Insights, the custom publishing division of MIT Technology Review. We were founded in 1899 at the Massachusetts Institute of Technology, and you can find us in print, on the web and at events each year around the world. For more information about us and the show, please check out our website at technologyreview.com.  

This show is available wherever you get your podcasts. And if you enjoyed this episode, we hope you’ll take a moment to rate and review us. Business Lab is a production of MIT Technology Review and this episode was produced by Giro Studios. Thanks for listening. 

This content was produced by Insights, the custom content arm of MIT Technology Review. It was not written by MIT Technology Review’s editorial staff. It was researched, designed, and written by human writers, editors, analysts, and illustrators. This includes the writing of surveys and collection of data for surveys. AI tools that may have been used were limited to secondary production processes that passed thorough human review.

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology.

Introducing: the AI Hype Correction package

AI is going to reproduce human intelligence. AI will eliminate disease. AI is the single biggest, most important invention in human history. You’ve likely heard it all—but probably none of these things are true.

AI is changing our world, but we don’t yet know the real winners, or how this will all shake out.

After a few years of out-of-control hype, people are now starting to re-calibrate what AI is, what it can do, and how we should think about its ultimate impact.

Here, at the end of 2025, we’re starting the post-hype phase. This new package of stories, called Hype Correction, is a way to reset expectations—a critical look at where we are, what AI makes possible, and where we go next.

Here’s a sneak peek at what you can expect:

+ An introduction to four ways of thinking about the great AI hype correction of 2025.

+  While it’s safe to say we’re definitely in an AI bubble right now, what’s less clear is what it really looks like—and what comes after it pops. Read the full story.

+ Why OpenAI’s Sam Altman can be traced back to so many of the more outlandish proclamations about AI doing the rounds these days. Read the full story.

+ It’s a weird time to be an AI doomer. But they’re not giving up.

+ AI coding is now everywhere—but despite the billions of dollars being poured into improving AI models’ coding abilities, not everyone is convinced. Read the full story.

+ If we really want to start finding new kinds of materials faster, AI materials discovery needs to make it out of the lab and move into the real world. Read the full story.

+ Why reports of AI’s potential to replace trained human lawyers are greatly exaggerated.

+ Dr. Margaret Mitchell, chief ethics scientist at AI startup Hugging Face, explains why the generative AI hype train is distracting us from what AI actually is and what it can—and crucially, cannot—do. Read the full story.

The must-reads

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

1 iRobot has filed for bankruptcy
The Roomba maker is considering handing over control to its main Chinese supplier. (Bloomberg $)
+ A proposed Amazon acquisition fell through close to two years ago. (FT $)
+ How the company lost its way. (TechCrunch)
+ A Roomba recorded a woman on the toilet. How did screenshots end up on Facebook? (MIT Technology Review)

2 Meta’s 2025 has been a total rollercoaster ride
From its controversial AI team to Mark Zuckerberg’s newfound appreciation for masculine energy. (Insider $)

3 The Trump administration is giving the crypto industry a much easier ride
It’s dismissed crypto lawsuits involving many firms with financial ties to Trump. (NYT $)
+ Celebrities are feeling emboldened to flog crypto once again. (The Guardian)
+ A bitcoin investor wants to set up a crypto libertarian community in the Caribbean. (FT $)

4 There’s a new weight-loss drug in town
And people are already taking it, even though it’s unapproved. (Wired $)
+ What we still don’t know about weight-loss drugs. (MIT Technology Review)

5 Chinese billionaires are having dozens of US-born surrogate babies
An entire industry has sprung up to support them. (WSJ $)
+ A controversial Chinese CRISPR scientist is still hopeful about embryo gene editing. (MIT Technology Review)

6 Trump’s “big beautiful bill” funding hinges on states integrating AI into healthcare
Experts fear it’ll be used as a cost-cutting measure, even if it doesn’t work. (The Guardian)
+ Artificial intelligence is infiltrating health care. We shouldn’t let it make all the decisions. (MIT Technology Review)

7 Extreme rainfall is wreaking havoc in the desert
Oman and the UAE are unaccustomed to increasingly common torrential downpours. (WP $)

8 Data centers are being built in countries that are too hot for them
Which makes it a lot harder to cool them sufficiently. (Rest of World)

9 Why AI image generators are getting deliberately worse
Their makers are pursuing realism—not that overly polished, Uncanny Valley look. (The Verge)
+ Inside the AI attention economy wars. (NY Mag $)

10 How a tiny Swedish city became a major video game hub
Skövde has formed an unlikely community of cutting-edge developers. (The Guardian)
+ Google DeepMind is using Gemini to train agents inside one of Skövde’s biggest franchises. (MIT Technology Review)

Quote of the day

“They don’t care about the games. They don’t care about the art. They just want their money.”

—Anna C Webster, chair of the freelancing committee of the United Videogame Workers union, tells the Guardian why their members are protesting the prestigious 2025 Game Awards in the wake of major layoffs.

One more thing

Recapturing early internet whimsy with HTML

Websites weren’t always slick digital experiences.

There was a time when surfing the web involved opening tabs that played music against your will and sifting through walls of text on a colored background. In the 2000s, before Squarespace and social media, websites were manifestations of individuality—built from scratch using HTML, by users who had some knowledge of code.

Scattered across the web are communities of programmers working to revive this seemingly outdated approach. And the movement is anything but a superficial appeal to retro aesthetics—it’s about celebrating the human touch in digital experiences. Read the full story.

—Tiffany Ng

We can still have nice things

A place for comfort, fun and distraction to brighten up your day. (Got any ideas? Drop me a line or skeet ’em at me.)

+  Here’s how a bit of math can help you wrap your presents much more neatly this year.
+ It seems that humans mastered making fire way, way earlier than we realized.
+ The Arab-owned cafes opening up across the US sound warm and welcoming.
+ How to give a gift the recipient will still be using and loving for decades to come.

Depending who you ask, AI-powered coding is either giving software developers an unprecedented productivity boost or churning out masses of poorly designed code that saps their attention and sets software projects up for serious long term-maintenance problems.

The problem is right now, it’s not easy to know which is true.

As tech giants pour billions into large language models (LLMs), coding has been touted as the technology’s killer app. Both Microsoft CEO Satya Nadella and Google CEO Sundar Pichai have claimed that around a quarter of their companies’ code is now AI-generated. And in March, Anthropic’s CEO, Dario Amodei, predicted that within six months 90% of all code would be written by AI. It’s an appealing and obvious use case. Code is a form of language, we need lots of it, and it’s expensive to produce manually. It’s also easy to tell if it works—run a program and it’s immediately evident whether it’s functional.

---

This story is part of MIT Technology Review’s Hype Correction package, a series that resets expectations about what AI is, what it makes possible, and where we go next.

---

Executives enamored with the potential to break through human bottlenecks are pushing engineers to lean into an AI-powered future. But after speaking to more than 30 developers, technology executives, analysts, and researchers, MIT Technology Review found that the picture is not as straightforward as it might seem.  

For some developers on the front lines, initial enthusiasm is waning as they bump up against the technology’s limitations. And as a growing body of research suggests that the claimed productivity gains may be illusory, some are questioning whether the emperor is wearing any clothes.

The pace of progress is complicating the picture, though. A steady drumbeat of new model releases mean these tools’ capabilities and quirks are constantly evolving. And their utility often depends on the tasks they are applied to and the organizational structures built around them. All of this leaves developers navigating confusing gaps between expectation and reality. 

Is it the best of times or the worst of times (to channel Dickens) for AI coding? Maybe both.

A fast-moving field

It’s hard to avoid AI coding tools these days. There are a dizzying array of products available, both from model developers like Anthropic, OpenAI, and Google and from companies like Cursor and Windsurf, which wrap these models in polished code-editing software. And according to Stack Overflow’s 2025 Developer Survey, they’re being adopted rapidly, with 65% of developers now using them at least weekly.

AI coding tools first emerged around 2016 but were supercharged with the arrival of LLMs. Early versions functioned as little more than autocomplete for programmers, suggesting what to type next. Today they can analyze entire code bases, edit across files, fix bugs, and even generate documentation explaining how the code works. All this is guided through natural-language prompts via a chat interface.

“Agents”—autonomous LLM-powered coding tools that can take a high-level plan and build entire programs independently—represent the latest frontier in AI coding. This leap was enabled by the latest reasoning models, which can tackle complex problems step by step and, crucially, access external tools to complete tasks. “This is how the model is able to code, as opposed to just talk about coding,” says Boris Cherny, head of Claude Code, Anthropic’s coding agent.

These agents have made impressive progress on software engineering benchmarks—standardized tests that measure model performance. When OpenAI introduced the SWE-bench Verified benchmark in August 2024, offering a way to evaluate agents’ success at fixing real bugs in open-source repositories, the top model solved just 33% of issues. A year later, leading models consistently score above 70%. 

In February, Andrej Karpathy, a founding member of OpenAI and former director of AI at Tesla, coined the term “vibe coding”—meaning an approach where people describe software in natural language and let AI write, refine, and debug the code. Social media abounds with developers who have bought into this vision, claiming massive productivity boosts.

But while some developers and companies report such productivity gains, the hard evidence is more mixed. Early studies from GitHub, Google, and Microsoft—all vendors of AI tools—found developers completing tasks 20% to 55% faster. But a September report from the consultancy Bain & Company described real-world savings as “unremarkable.”

Data from the developer analytics firm GitClear shows that most engineers are producing roughly 10% more durable code—code that isn’t deleted or rewritten within weeks—since 2022, likely thanks to AI. But that gain has come with sharp declines in several measures of code quality. Stack Overflow’s survey also found trust and positive sentiment toward AI tools falling significantly for the first time. And most provocatively, a July study by the nonprofit research organization Model Evaluation & Threat Research (METR) showed that while experienced developers believed AI made them 20% faster, objective tests showed they were actually 19% slower.

Growing disillusionment

For Mike Judge, principal developer at the software consultancy Substantial, the METR study struck a nerve. He was an enthusiastic early adopter of AI tools, but over time he grew frustrated with their limitations and the modest boost they brought to his productivity. “I was complaining to people because I was like, ‘It’s helping me but I can’t figure out how to make it really help me a lot,’” he says. “I kept feeling like the AI was really dumb, but maybe I could trick it into being smart if I found the right magic incantation.”

When asked by a friend, Judge had estimated the tools were providing a roughly 25% speedup. So when he saw similar estimates attributed to developers in the METR study he decided to test his own. For six weeks, he guessed how long a task would take, flipped a coin to decide whether to use AI or code manually, and timed himself. To his surprise, AI slowed him down by an median of 21%—mirroring the METR results.

This got Judge crunching the numbers. If these tools were really speeding developers up, he reasoned, you should see a massive boom in new apps, website registrations, video games, and projects on GitHub. He spent hours and several hundred dollars analyzing all the publicly available data and found flat lines everywhere.

“Shouldn’t this be going up and to the right?” says Judge. “Where’s the hockey stick on any of these graphs? I thought everybody was so extraordinarily productive.” The obvious conclusion, he says, is that AI tools provide little productivity boost for most developers. 

Developers interviewed by MIT Technology Review generally agree on where AI tools excel: producing “boilerplate code” (reusable chunks of code repeated in multiple places with little modification), writing tests, fixing bugs, and explaining unfamiliar code to new developers. Several noted that AI helps overcome the “blank page problem” by offering an imperfect first stab to get a developer’s creative juices flowing. It can also let nontechnical colleagues quickly prototype software features, easing the load on already overworked engineers.

These tasks can be tedious, and developers are typically  glad to hand them off. But they represent only a small part of an experienced engineer’s workload. For the more complex problems where engineers really earn their bread, many developers told MIT Technology Review, the tools face significant hurdles.

Perhaps the biggest problem is that LLMs can hold only a limited amount of information in their “context window”—essentially their working memory. This means they struggle to parse large code bases and are prone to forgetting what they’re doing on longer tasks. “It gets really nearsighted—it’ll only look at the thing that’s right in front of it,” says Judge. “And if you tell it to do a dozen things, it’ll do 11 of them and just forget that last one.”

LLMs’ myopia can lead to headaches for human coders. While an LLM-generated response to a problem may work in isolation, software is made up of hundreds of interconnected modules. If these aren’t built with consideration for other parts of the software, it can quickly lead to a tangled, inconsistent code base that’s hard for humans to parse and, more important, to maintain.

Developers have traditionally addressed this by following conventions—loosely defined coding guidelines that differ widely between projects and teams. “AI has this overwhelming tendency to not understand what the existing conventions are within a repository,” says Bill Harding, the CEO of GitClear. “And so it is very likely to come up with its own slightly different version of how to solve a problem.”

The models also just get things wrong. Like all LLMs, coding models are prone to “hallucinating”—it’s an issue built into how they work. But because the code they output looks so polished, errors can be difficult to detect, says James Liu, director of software engineering at the advertising technology company Mediaocean. Put all these flaws together, and using these tools can feel a lot like pulling a lever on a one-armed bandit. “Some projects you get a 20x improvement in terms of speed or efficiency,” says Liu. “On other things, it just falls flat on its face, and you spend all this time trying to coax it into granting you the wish that you wanted and it’s just not going to.”

Judge suspects this is why engineers often overestimate productivity gains. “You remember the jackpots. You don’t remember sitting there plugging tokens into the slot machine for two hours,” he says.

And it can be particularly pernicious if the developer is unfamiliar with the task. Judge remembers getting AI to help set up a Microsoft cloud service called an Azure Functions, which he’d never used before. He thought it would take about two hours, but nine hours later he threw in the towel. “It kept leading me down these rabbit holes and I didn’t know enough about the topic to be able to tell it ‘Hey, this is nonsensical,’” he says.

The debt begins to mount up

Developers constantly make trade-offs between speed of development and the maintainability of their code—creating what’s known as “technical debt,” says Geoffrey G. Parker, professor of engineering innovation at Dartmouth College. Each shortcut adds complexity and makes the code base harder to manage, accruing “interest” that must eventually be repaid by restructuring the code. As this debt piles up, adding new features and maintaining the software becomes slower and more difficult.

Accumulating technical debt is inevitable in most projects, but AI tools make it much easier for time-pressured engineers to cut corners, says GitClear’s Harding. And GitClear’s data suggests this is happening at scale. Since 2020, the company has seen a significant rise in the amount of copy-pasted code—an indicator that developers are reusing more code snippets, most likely based on AI suggestions—and an even bigger decline in the amount of code moved from one place to another, which happens when developers clean up their code base.

And as models improve, the code they produce is becoming increasingly verbose and complex, says Tariq Shaukat, CEO of Sonar, which makes tools for checking code quality. This is driving down the number of obvious bugs and security vulnerabilities, he says, but at the cost of increasing the number of “code smells”—harder-to-pinpoint flaws that lead to maintenance problems and technical debt. 

Recent research by Sonar found that these make up more than 90% of the issues found in code generated by leading AI models. “Issues that are easy to spot are disappearing, and what’s left are much more complex issues that take a while to find,” says Shaukat. “That’s what worries us about this space at the moment. You’re almost being lulled into a false sense of security.”

If AI tools make it increasingly difficult to maintain code, that could have significant security implications, says Jessica Ji, a security researcher at Georgetown University. “The harder it is to update things and fix things, the more likely a code base or any given chunk of code is to become insecure over time,” says Ji.

There are also more specific security concerns, she says. Researchers have discovered a worrying class of hallucinations where models reference nonexistent software packages in their code. Attackers can exploit this by creating packages with those names that harbor vulnerabilities, which the model or developer may then unwittingly incorporate into software. 

LLMs are also vulnerable to “data-poisoning attacks,” where hackers seed the publicly available data sets models train on with data that alters the model’s behavior in undesirable ways, such as generating insecure code when triggered by specific phrases. In October, research by Anthropic found that as few as 250 malicious documents can introduce this kind of back door into an LLM regardless of its size.

The converted

Despite these issues, though, there’s probably no turning back. “Odds are that writing every line of code on a keyboard by hand—those days are quickly slipping behind us,” says Kyle Daigle, chief operating officer at the Microsoft-owned code-hosting platform GitHub, which produces a popular AI-powered tool called Copilot (not to be confused with the Microsoft product of the same name).

The Stack Overflow report found that despite growing distrust in the technology, usage has increased rapidly and consistently over the past three years. Erin Yepis, a senior analyst at Stack Overflow, says this suggests that engineers are taking advantage of the tools with a clear-eyed view of the risks. The report also found that frequent users tend to be more enthusiastic and more than half of developers are not using the latest coding agents, perhaps explaining why many remain underwhelmed by the technology.

Those latest tools can be a revelation. Trevor Dilley, CTO at the software development agency Twenty20 Ideas, says he had found some value in AI editors’ autocomplete functions, but when he tried anything more complex it would “fail catastrophically.” Then in March, while on vacation with his family, he set the newly released Claude Code to work on one of his hobby projects. It completed a four-hour task in two minutes, and the code was better than what he would have written.

“I was like, Whoa,” he says. “That, for me, was the moment, really. There’s no going back from here.” Dilley has since cofounded a startup called DevSwarm, which is creating software that can marshal multiple agents to work in parallel on a piece of software.

The challenge, says Armin Ronacher, a prominent open-source developer, is that the learning curve for these tools is shallow but long. Until March he’d remained unimpressed by AI tools, but after leaving his job at the software company Sentry in April to launch a startup, he started experimenting with agents. “I basically spent a lot of months doing nothing but this,” he says. “Now, 90% of the code that I write is AI-generated.”

Getting to that point involved extensive trial and error, to figure out which problems tend to trip the tools up and which they can handle efficiently. Today’s models can tackle most coding tasks with the right guardrails, says Ronacher, but these can be very task and project specific.

To get the most out of these tools, developers must surrender control over individual lines of code and focus on the overall software architecture, says Nico Westerdale, chief technology officer at the veterinary staffing company IndeVets. He recently built a data science platform 100,000 lines of code long almost exclusively by prompting models rather than writing the code himself.

Westerdale’s process starts with an extended conversation with the modelagent to develop a detailed plan for what to build and how. He then guides it through each step. It rarely gets things right on the first try and needs constant wrangling, but if you force it to stick to well-defined design patterns, the models can produce high-quality, easily maintainable code, says Westerdale. He reviews every line, and the code is as good as anything he’s ever produced, he says: “I’ve just found it absolutely revolutionary,. It’s also frustrating, difficult, a different way of thinking, and we’re only just getting used to it.”

But while individual developers are learning how to use these tools effectively, getting consistent results across a large engineering team is significantly harder. AI tools amplify both the good and bad aspects of your engineering culture, says Ryan J. Salva, senior director of product management at Google. With strong processes, clear coding patterns, and well-defined best practices, these tools can shine. 

But if your development process is disorganized, they’ll only magnify the problems. It’s also essential to codify that institutional knowledge so the models can draw on it effectively. “A lot of work needs to be done to help build up context and get the tribal knowledge out of our heads,” he says.

The cryptocurrency exchange Coinbase has been vocal about its adoption of AI tools. CEO Brian Armstrong made headlines in August when he revealed that the company had fired staff unwilling to adopt AI tools. But Coinbase’s head of platform, Rob Witoff, tells MIT Technology Review that while they’ve seen massive productivity gains in some areas, the impact has been patchy. For simpler tasks like restructuring the code base and writing tests, AI-powered workflows have achieved speedups of up to 90%. But gains are more modest for other tasks, and the disruption caused by overhauling existing processes often counteracts the increased coding speed, says Witoff.

One factor is that AI tools let junior developers produce far more code,. As in almost all engineering teams, this code has to be reviewed by others, normally more senior developers, to catch bugs and ensure it meets quality standards. But the sheer volume of code now being churned out i whichs quickly saturatinges the ability of midlevel staff to review changes. “This is the cycle we’re going through almost every month, where we automate a new thing lower down in the stack, which brings more pressure higher up in the stack,” he says. “Then we’re looking at applying automation to that higher-up piece.”

Developers also spend only 20% to 40% of their time coding, says Jue Wang, a partner at Bain, so even a significant speedup there often translates to more modest overall gains. Developers spend the rest of their time analyzing software problems and dealing with customer feedback, product strategy, and administrative tasks. To get significant efficiency boosts, companies may need to apply generative AI to all these other processes too, says Jue, and that is still in the works.

Rapid evolution

Programming with agents is a dramatic departure from previous working practices, though, so it’s not surprising companies are facing some teething issues. These are also very new products that are changing by the day. “Every couple months the model improves, and there’s a big step change in the model’s coding capabilities and you have to get recalibrated,” says Anthropic’s Cherny.

For example, in June Anthropic introduced a built-in planning mode to Claude; it has since been replicated by other providers. In October, the company also enabled Claude to ask users questions when it needs more context or faces multiple possible solutions, which Cherny says helps it avoid the tendency to simply assume which path is the best way forward.

Most significant, Anthropic has added features that make Claude better at managing its own context. When it nears the limits of its working memory, it summarizes key details and uses them to start a new context window, effectively giving it an “infinite” one, says Cherny. Claude can also invoke sub-agents to work on smaller tasks, so it no longer has to hold all aspects of the project in its own head. The company claims that its latest model, Claude 4.5 Sonnet, can now code autonomously for more than 30 hours without major performance degradation.

Novel approaches to software development could also sidestep coding agents’ other flaws. MIT professor Max Tegmark has introduced something he calls “vericoding,” which could allow agents to produce entirely bug-free code from a natural-language description. It builds on an approach known as “formal verification,” where developers create a mathematical model of their software that can prove incontrovertibly that it functions correctly. This approach is used in high-stakes areas like flight-control systems and cryptographic libraries, but it remains costly and time-consuming, limiting its broader use.

Rapid improvements in LLMs’ mathematical capabilities have opened up the tantalizing possibility of models that produce not only software but the mathematical proof that it’s bug free, says Tegmark. “You just give the specification, and the AI comes back with provably correct code,” he says. “You don’t have to touch the code. You don’t even have to ever look at the code.”

When tested on about 2,000 vericoding problems in Dafny—a language designed for formal verification—the best LLMs solved over 60%, according to non-peer-reviewed research by Tegmark’s group. This was achieved with off-the-shelf LLMs, and Tegmark expects that training specifically for vericoding could improve scores rapidly.

And counterintuitively, Tthe speed at which AI generates code could actuallylso ease maintainability concerns. Alex Worden, principal engineer at the business software giant Intuit, notes that maintenance is often difficult because engineers reuse components across projects, creating a tangle of dependencies where one change triggers cascading effects across the code base. Reusing code used to save developers time, but in a world where AI can produce hundreds of lines of code in seconds, that imperative has gone, says Worden.

Instead, he advocates for “disposable code,” where each component is generated independently by AI without regard for whether it follows design patterns or conventions. They are then connected via APIs—sets of rules that let components request information or services from each other. Each component’s inner workings are not dependent on other parts of the code base, making it possible to rip them out and replace them without wider impact, says Worden. 

“The industry is still concerned about humans maintaining AI-generated code,” he says. “I question how long humans will look at or care about code.”

A narrowing talent pipeline

For the foreseeable future, though, humans will still need to understand and maintain the code that underpins their projects. And one of the most pernicious side effects of AI tools may be a shrinking pool of people capable of doing so. 

Early evidence suggests that fears around the job-destroying effects of AI may be justified. A recent Stanford University study found that employment among software developers aged 22 to 25 fell nearly 20% between 2022 and 2025, coinciding with the rise of AI-powered coding tools.

Experienced developers could face difficulties too. Luciano Nooijen, an engineer at the video-game infrastructure developer Companion Group, used AI tools heavily in his day job, where they were provided for free. But when he began a side project without access to those tools, he found himself struggling with tasks that previously came naturally. “I was feeling so stupid because things that used to be instinct became manual, sometimes even cumbersome,” says Nooijen.

Just as athletes still perform basic drills, he thinks the only way to maintain an instinct for coding is to regularly practice the grunt work. That’s why he’s largely abandoned AI tools, though he admits that deeper motivations are also at play. 

Part of the reason Nooijen and other developers MIT Technology Review spoke to are pushing back against AI tools is a sense that they are hollowing out the parts of their jobs that they love. “I got into software engineering because I like working with computers. I like making machines do things that I want,” Nooijen says. “It’s just not fun sitting there with my work being done for me.”

Each time you’ve heard a borderline outlandish idea of what AI will be capable of, it often turns out that Sam Altman was, if not the first to articulate it, at least the most persuasive and influential voice behind it. 

For more than a decade he has been known in Silicon Valley as a world-class fundraiser and persuader. OpenAI’s early releases around 2020 set the stage for a mania around large language models, and the launch of ChatGPT in November 2022 granted Altman a world stage on which to present his new thesis: that these models mirror human intelligence and could swing the doors open to a healthier and wealthier techno-utopia.

---

This story is part of MIT Technology Review’s Hype Correction package, a series that resets expectations about what AI is, what it makes possible, and where we go next.

---

Throughout, Altman’s words have set the agenda. He has framed a prospective superintelligent AI as either humanistic or catastrophic, depending on what effect he was hoping to create, what he was raising money for, or which tech giant seemed like his most formidable competitor at the moment. 

Examining Altman’s statements over the years reveals just how much his outlook has powered today’s AI boom. Even among Silicon Valley’s many hypesters, he’s been especially willing to speak about open questions—whether large language models contain the ingredients of human thought, whether language can also produce intelligence—as if they were already answered. 

In conclusion … 

Altman didn’t dupe the world. OpenAI has ushered in a genuine tech revolution, with increasingly impressive language models that have attracted millions of users. Even skeptics would concede that LLMs’ conversational ability is astonishing.

But Altman’s hype has always hinged less on today’s capabilities than on a philosophical tomorrow—an outlook that quite handily doubles as a case for more capital and friendlier regulation. Long before large language models existed, he was imagining an AI powerful enough to require wealth redistribution, just as he imagined humanity colonizing other planets. Again and again, promises of a destination—abundance, superintelligence, a healthier and wealthier world—have come first, and the evidence second. 

Even if LLMs eventually hit a wall, there’s little reason to think his faith in a techno-utopian future will falter. The vision was never really about the particulars of the current model anyway. 

It’s a weird time to be an AI doomer.

This small but influential community of researchers, scientists, and policy experts believes, in the simplest terms, that AI could get so good it could be bad—very, very bad—for humanity. Though many of these people would be more likely to describe themselves as advocates for AI safety than as literal doomsayers, they warn that AI poses an existential risk to humanity. They argue that absent more regulation, the industry could hurtle toward systems it can’t control. They commonly expect such systems to follow the creation of artificial general intelligence (AGI), a slippery concept generally understood as technology that can do whatever humans can do, and better. 

---

This story is part of MIT Technology Review’s Hype Correction package, a series that resets expectations about what AI is, what it makes possible, and where we go next.

---

Though this is far from a universally shared perspective in the AI field, the doomer crowd has had some notable success over the past several years: helping shape AI policy coming from the Biden administration, organizing prominent calls for international “red lines” to prevent AI risks, and getting a bigger (and more influential) megaphone as some of its adherents win science’s most prestigious awards.

But a number of developments over the past six months have put them on the back foot. Talk of an AI bubble has overwhelmed the discourse as tech companies continue to invest in multiple Manhattan Projects’ worth of data centers without any certainty that future demand will match what they’re building. 

And then there was the August release of OpenAI’s latest foundation model, GPT-5, which proved something of a letdown. Maybe that was inevitable, since it was the most hyped AI release of all time; OpenAI CEO Sam Altman had boasted that GPT-5 felt “like a PhD-level expert” in every topic and told the podcaster Theo Von that the model was so good, it had made him feel “useless relative to the AI.” 

Many expected GPT-5 to be a big step toward AGI, but whatever progress the model may have made was overshadowed by a string of technical bugs and the company’s mystifying, quickly reversed decision to shut off access to every old OpenAI model without warning. And while the new model achieved state-of-the-art benchmark scores, many people felt, perhaps unfairly, that in day-to-day use GPT-5 was a step backward. 

All this would seem to threaten some of the very foundations of the doomers’ case. In turn, a competing camp of AI accelerationists, who fear AI is actually not moving fast enough and that the industry is constantly at risk of being smothered by overregulation, is seeing a fresh chance to change how we approach AI safety (or, maybe more accurately, how we don’t). 

This is particularly true of the industry types who’ve decamped to Washington: “The Doomer narratives were wrong,” declared David Sacks, the longtime venture capitalist turned Trump administration AI czar. “This notion of imminent AGI has been a distraction and harmful and now effectively proven wrong,” echoed the White House’s senior policy advisor for AI and tech investor Sriram Krishnan. (Sacks and Krishnan did not reply to requests for comment.) 

(There is, of course, another camp in the AI safety debate: the group of researchers and advocates commonly associated with the label “AI ethics.” Though they also favor regulation, they tend to think the speed of AI progress has been overstated and have often written off AGI as a sci-fi story or a scam that distracts us from the technology’s immediate threats. But any potential doomer demise wouldn’t exactly give them the same opening the accelerationists are seeing.)

So where does this leave the doomers? As part of our Hype Correction package, we decided to ask some of the movement’s biggest names to see if the recent setbacks and general vibe shift had altered their views. Are they frustrated that policymakers no longer seem to heed their threats? Are they quietly adjusting their timelines for the apocalypse? 

Recent interviews with 20 people who study or advocate AI safety and governance—including Nobel Prize winner Geoffrey Hinton, Turing Prize winner Yoshua Bengio, and high-profile experts like former OpenAI board member Helen Toner—reveal that rather than feeling chastened or lost in the wilderness, they’re still deeply committed to their cause, believing that AGI remains not just possible but incredibly dangerous.

At the same time, they seem to be grappling with a near contradiction. While they’re somewhat relieved that recent developments suggest AGI is further out than they previously thought (“Thank God we have more time,” says AI researcher Jeffrey Ladish), they also feel angry that people in power are not taking them seriously enough (Daniel Kokotajlo, lead author of a cautionary forecast called “AI 2027,” calls the Sacks and Krishnan tweets “deranged and/or dishonest”). 

Broadly speaking, these experts see the talk of an AI bubble as no more than a speed bump, and disappointment in GPT-5 as more distracting than illuminating. They still generally favor more robust regulation and worry that progress on policy—the implementation of the EU AI Act; the passage of the first major American AI safety bill, California’s SB 53; and new interest in AGI risk from some members of Congress—has become vulnerable as Washington overreacts to what doomers see as short-term failures to live up to the hype. 

Some were also eager to correct what they see as the most persistent misconceptions about the doomer world. Though their critics routinely mock them for predicting that AGI is right around the corner, they claim that’s never been an essential part of their case: It “isn’t about imminence,” says Berkeley professor Stuart Russell, the author of Human Compatible: Artificial Intelligence and the Problem of Control. Most people I spoke with say their timelines to dangerous systems have actually lengthened slightly in the last year—an important change given how quickly the policy and technical landscapes can shift. 

“If someone said there’s a four-mile-diameter asteroid that’s going to hit the Earth in 2067, we wouldn’t say, ‘Remind me in 2066 and we’ll think about it.’”

Many of them, in fact, emphasize the importance of changing timelines. And even if they are just a tad longer now, Toner tells me that one big-picture story of the ChatGPT era is the dramatic compression of these estimates across the AI world. For a long while, she says, AGI was expected in many decades. Now, for the most part, the predicted arrival is sometime in the next few years to 20 years. So even if we have a little bit more time, she (and many of her peers) continue to see AI safety as incredibly, vitally urgent. She tells me that if AGI were possible anytime in even the next 30 years, “It’s a huge fucking deal. We should have a lot of people working on this.”

So despite the precarious moment doomers find themselves in, their bottom line remains that no matter when AGI is coming (and, again, they say it’s very likely coming), the world is far from ready. 

Maybe you agree. Or maybe you may think this future is far from guaranteed. Or that it’s the stuff of science fiction. You may even think AGI is a great big conspiracy theory. You’re not alone, of course—this topic is polarizing. But whatever you think about the doomer mindset, there’s no getting around the fact that certain people in this world have a lot of influence. So here are some of the most prominent people in the space, reflecting on this moment in their own words. 

Interviews have been edited and condensed for length and clarity. 

---

The Nobel laureate who’s not sure what’s coming

Geoffrey Hinton, winner of the Turing Award and the Nobel Prize in physics for pioneering deep learning

The biggest change in the last few years is that there are people who are hard to dismiss who are saying this stuff is dangerous. Like, [former Google CEO] Eric Schmidt, for example, really recognized this stuff could be really dangerous. He and I were in China recently talking to someone on the Politburo, the party secretary of Shanghai, to make sure he really understood—and he did. I think in China, the leadership understands AI and its dangers much better because many of them are engineers.

I’ve been focused on the longer-term threat: When AIs get more intelligent than us, can we really expect that humans will remain in control or even relevant? But I don’t think anything is inevitable. There’s huge uncertainty on everything. We’ve never been here before. Anybody who’s confident they know what’s going to happen seems silly to me. I think this is very unlikely but maybe it’ll turn out that all the people saying AI is way overhyped are correct. Maybe it’ll turn out that we can’t get much further than the current chatbots—we hit a wall due to limited data. I don’t believe that. I think that’s unlikely, but it’s possible. 

I also don’t believe people like Eliezer Yudkowsky, who say if anybody builds it, we’re all going to die. We don’t know that. 

But if you go on the balance of the evidence, I think it’s fair to say that most experts who know a lot about AI believe it’s very probable that we’ll have superintelligence within the next 20 years. [Google DeepMind CEO] Demis Hassabis says maybe 10 years. Even [prominent AI skeptic] Gary Marcus would probably say, “Well, if you guys make a hybrid system with good old-fashioned symbolic logic … maybe that’ll be superintelligent.” [Editor’s note: In September, Marcus predicted AGI would arrive between 2033 and 2040.]

And I don’t think anybody believes progress will stall at AGI. I think more or less everybody believes a few years after AGI, we’ll have superintelligence, because the AGI will be better than us at building AI.

So while I think it’s clear that the winds are getting more difficult, simultaneously, people are putting in many more resources [into developing advanced AI]. I think progress will continue just because there’s many more resources going in.

The deep learning pioneer who wishes he’d seen the risks sooner

Yoshua Bengio, winner of the Turing Award, chair of the International AI Safety Report, and founder of LawZero

Some people thought that GPT-5 meant we had hit a wall, but that isn’t quite what you see in the scientific data and trends.

There have been people overselling the idea that AGI is tomorrow morning, which commercially could make sense. But if you look at the various benchmarks, GPT-5 is just where you would expect the models at that point in time to be. By the way, it’s not just GPT-5, it’s Claude and Google models, too. In some areas where AI systems weren’t very good, like Humanity’s Last Exam or FrontierMath, they’re getting much better scores now than they were at the beginning of the year.

At the same time, the overall landscape for AI governance and safety is not good. There’s a strong force pushing against regulation. It’s like climate change. We can put our head in the sand and hope it’s going to be fine, but it doesn’t really deal with the issue.

The biggest disconnect with policymakers is a misunderstanding of the scale of change that is likely to happen if the trend of AI progress continues. A lot of people in business and governments simply think of AI as just another technology that’s going to be economically very powerful. They don’t understand how much it might change the world if trends continue, and we approach human-level AI. 

Like many people, I had been blinding myself to the potential risks to some extent. I should have seen it coming much earlier. But it’s human. You’re excited about your work and you want to see the good side of it. That makes us a little bit biased in not really paying attention to the bad things that could happen.

Even a small chance—like 1% or 0.1%—of creating an accident where billions of people die is not acceptable. 

The AI veteran who believes AI is progressing—but not fast enough to prevent the bubble from bursting

Stuart Russell, distinguished professor of computer science, University of California, Berkeley, and author of Human Compatible

I hope the idea that talking about existential risk makes you a “doomer” or is “science fiction” comes to be seen as fringe, given that most leading AI researchers and most leading AI CEOs take it seriously. 

There have been claims that AI could never pass a Turing test, or you could never have a system that uses natural language fluently, or one that could parallel-park a car. All these claims just end up getting disproved by progress.

People are spending trillions of dollars to make superhuman AI happen. I think they need some new ideas, but there’s a significant chance they will come up with them, because many significant new ideas have happened in the last few years. 

My fairly consistent estimate for the last 12 months has been that there’s a 75% chance that those breakthroughs are not going to happen in time to rescue the industry from the bursting of the bubble. Because the investments are consistent with a prediction that we’re going to have much better AI that will deliver much more value to real customers. But if those predictions don’t come true, then there’ll be a lot of blood on the floor in the stock markets.

However, the safety case isn’t about imminence. It’s about the fact that we still don’t have a solution to the control problem. If someone said there’s a four-mile-diameter asteroid that’s going to hit the Earth in 2067, we wouldn’t say, “Remind me in 2066 and we’ll think about it.” We don’t know how long it takes to develop the technology needed to control superintelligent AI.

Looking at precedents, the acceptable level of risk for a nuclear plant melting down is about one in a million per year. Extinction is much worse than that. So maybe set the acceptable risk at one in a billion. But the companies are saying it’s something like one in five. They don’t know how to make it acceptable. And that’s a problem.

The professor trying to set the narrative straight on AI safety

David Krueger, assistant professor in machine learning at the University of Montreal and Yoshua Bengio’s Mila Institute, and founder of Evitable

I think people definitely overcorrected in their response to GPT-5. But there was hype. My recollection was that there were multiple statements from CEOs at various levels of explicitness who basically said that by the end of 2025, we’re going to have an automated drop-in replacement remote worker. But it seems like it’s been underwhelming, with agents just not really being there yet.

I’ve been surprised how much these narratives predicting AGI in 2027 capture the public attention. When 2027 comes around, if things still look pretty normal, I think people are going to feel like the whole worldview has been falsified. And it’s really annoying how often when I’m talking to people about AI safety, they assume that I think we have really short timelines to dangerous systems, or that I think LLMs or deep learning are going to give us AGI. They ascribe all these extra assumptions to me that aren’t necessary to make the case. 

I’d expect we need decades for the international coordination problem. So even if dangerous AI is decades off, it’s already urgent. That point seems really lost on a lot of people. There’s this idea of “Let’s wait until we have a really dangerous system and then start governing it.” Man, that is way too late.

I still think people in the safety community tend to work behind the scenes, with people in power, not really with civil society. It gives ammunition to people who say it’s all just a scam or insider lobbying. That’s not to say that there’s no truth to these narratives, but the underlying risk is still real. We need more public awareness and a broad base of support to have an effective response.

If you actually believe there’s a 10% chance of doom in the next 10 years—which I think a reasonable person should, if they take a close look—then the first thing you think is: “Why are we doing this? This is crazy.” That’s just a very reasonable response once you buy the premise.

The governance expert worried about AI safety’s credibility

Helen Toner, acting executive director of Georgetown University’s Center for Security and Emerging Technology and former OpenAI board member

When I got into the space, AI safety was more of a set of philosophical ideas. Today, it’s a thriving set of subfields of machine learning, filling in the gulf between some of the more “out there” concerns about AI scheming, deception, or power-seeking and real concrete systems we can test and play with. 

“I worry that some aggressive AGI timeline estimates from some AI safety people are setting them up for a boy-who-cried-wolf moment.”

AI governance is improving slowly. If we have lots of time to adapt and governance can keep improving slowly, I feel not bad. If we don’t have much time, then we’re probably moving too slow.

I think GPT-5 is generally seen as a disappointment in DC. There’s a pretty polarized conversation around: Are we going to have AGI and superintelligence in the next few years? Or is AI actually just totally all hype and useless and a bubble? The pendulum had maybe swung too far toward “We’re going to have super-capable systems very, very soon.” And so now it’s swinging back toward “It’s all hype.”

I worry that some aggressive AGI timeline estimates from some AI safety people are setting them up for a boy-who-cried-wolf moment. When the predictions about AGI coming in 2027 don’t come true, people will say, “Look at all these people who made fools of themselves. You should never listen to them again.” That’s not the intellectually honest response, if maybe they later changed their mind, or their take was that they only thought it was 20 percent likely and they thought that was still worth paying attention to. I think that shouldn’t be disqualifying for people to listen to you later, but I do worry it will be a big credibility hit. And that’s applying to people who are very concerned about AI safety and never said anything about very short timelines.

The AI security researcher who now believes AGI is further out—and is grateful

Jeffrey Ladish, executive director at Palisade Research

In the last year, two big things updated my AGI timelines. 

First, the lack of high-quality data turned out to be a bigger problem than I expected. 

Second, the first “reasoning” model, OpenAI’s o1 in September 2024, showed reinforcement learning scaling was more effective than I thought it would be. And then months later, you see the o1 to o3 scale-up and you see pretty crazy impressive performance in math and coding and science—domains where it’s easier to sort of verify the results. But while we’re seeing continued progress, it could have been much faster.

All of this bumps up my median estimate to the start of fully automated AI research and development from three years to maybe five or six years. But those are kind of made up numbers. It’s hard. I want to caveat all this with, like, “Man, it’s just really hard to do forecasting here.”

Thank God we have more time. We have a possibly very brief window of opportunity to really try to understand these systems before they are capable and strategic enough to pose a real threat to our ability to control them.

But it’s scary to see people think that we’re not making progress anymore when that’s clearly not true. I just know it’s not true because I use the models. One of the downsides of the way AI is progressing is that how fast it’s moving is becoming less legible to normal people. 

Now, this is not true in some domains—like, look at Sora 2. It is so obvious to anyone who looks at it that Sora 2 is vastly better than what came before. But if you ask GPT-4 and GPT-5 why the sky is blue, they’ll give you basically the same answer. It is the correct answer. It’s already saturated the ability to tell you why the sky is blue. So the people who I expect to most understand AI progress right now are the people who are actually building with AIs or using AIs on very difficult scientific problems.

The AGI forecaster who saw the critics coming

Daniel Kokotajlo, executive director of the AI Futures Project; an OpenAI whistleblower; and lead author of “AI 2027,” a vivid scenario where—starting in 2027—AIs progress from “superhuman coders” to “wildly superintelligent” systems in the span of months

AI policy seems to be getting worse, like the “Pro-AI” super PAC [launched earlier this year by executives from OpenAI and Andreessen Horowitz to lobby for a deregulatory agenda], and the deranged and/or dishonest tweets from Sriram Krishnan and David Sacks. AI safety research is progressing at the usual pace, which is excitingly rapid compared to most fields, but slow compared to how fast it needs to be.

We said on the first page of “AI 2027” that our timelines were somewhat longer than 2027. So even when we launched AI 2027, we expected there to be a bunch of critics in 2028 triumphantly saying we’ve been discredited, like the tweets from Sacks and Krishnan. But we thought, and continue to think, that the intelligence explosion will probably happen sometime in the next five to 10 years, and that when it does, people will remember our scenario and realize it was closer to the truth than anything else available in 2025. 

Predicting the future is hard, but it’s valuable to try; people should aim to communicate their uncertainty about the future in a way that is specific and falsifiable. This is what we’ve done and very few others have done. Our critics mostly haven’t made predictions of their own and often exaggerate and mischaracterize our views. They say our timelines are shorter than they are or ever were, or they say we are more confident than we are or were.

I feel pretty good about having longer timelines to AGI. It feels like I just got a better prognosis from my doctor. The situation is still basically the same, though.

Garrison Lovely is a freelance journalist and the author of Obsolete, an online publication and forthcoming book on the discourse, economics, and geopolitics of the race to build machine superintelligence (out spring 2026). His writing on AI has appeared in the New York Times, Nature, Bloomberg, Time, the Guardian, The Verge, and elsewhere.

Some disillusionment was inevitable. When OpenAI released a free web app called ChatGPT in late 2022, it changed the course of an entire industry—and several world economies. Millions of people started talking to their computers, and their computers started talking back. We were enchanted, and we expected more.

We got it. Technology companies scrambled to stay ahead, putting out rival products that outdid one another with each new release: voice, images, video. With nonstop one-upmanship, AI companies have presented each new product drop as a major breakthrough, reinforcing a widespread faith that this technology would just keep getting better. Boosters told us that progress was exponential. They posted charts plotting how far we’d come since last year’s models: Look how the line goes up! Generative AI could do anything, it seemed.

Well, 2025 has been a year of reckoning. 

---

This story is part of MIT Technology Review’s Hype Correction package, a series that resets expectations about what AI is, what it makes possible, and where we go next.

---

For a start, the heads of the top AI companies made promises they couldn’t keep. They told us that generative AI would replace the white-collar workforce, bring about an age of abundance, make scientific discoveries, and help find new cures for disease. FOMO across the world’s economies, at least in the Global North, made CEOs tear up their playbooks and try to get in on the action.

That’s when the shine started to come off. Though the technology may have been billed as a universal multitool that could revamp outdated business processes and cut costs, a number of studies published this year suggest that firms are failing to make the AI pixie dust work its magic. Surveys and trackers from a range of sources, including the US Census Bureau and Stanford University, have found that business uptake of AI tools is stalling. And when the tools do get tried out, many projects stay stuck in the pilot stage. Without broad buy-in across the economy it is not clear how the big AI companies will ever recoup the incredible amounts they’ve already spent in this race. 

At the same time, updates to the core technology are no longer the step changes they once were.

The highest-profile example of this was the botched launch of GPT-5 in August. Here was OpenAI, the firm that had ignited (and to a large extent sustained) the current boom, set to release a brand-new generation of its technology. OpenAI had been hyping GPT-5 for months: “PhD-level expert in anything,” CEO Sam Altman crowed. On another occasion Altman posted, without comment, an image of the Death Star from Star Wars, which OpenAI stans took to be a symbol of ultimate power: Coming soon! Expectations were huge.

And yet, when it landed, GPT-5 seemed to be—more of the same? What followed was the biggest vibe shift since ChatGPT first appeared three years ago. “The era of boundary-breaking advancements is over,” Yannic Kilcher, an AI researcher and popular YouTuber, announced in a video posted two days after GPT-5 came out: “AGI is not coming. It seems very much that we’re in the Samsung Galaxy era of LLMs.”

A lot of people (me included) have made the analogy with phones. For a decade or so, smartphones were the most exciting consumer tech in the world. Today, new products drop from Apple or Samsung with little fanfare. While superfans pore over small upgrades, to most people this year’s iPhone now looks and feels a lot like last year’s iPhone. Is that where we are with generative AI? And is it a problem? Sure, smartphones have become the new normal. But they changed the way the world works, too.

To be clear, the last few years have been filled with genuine “Wow” moments, from the stunning leaps in the quality of video generation models to the problem-solving chops of so-called reasoning models to the world-class competition wins of the latest coding and math models. But this remarkable technology is only a few years old, and in many ways it is still experimental. Its successes come with big caveats.

Perhaps we need to readjust our expectations.

The big reset

Let’s be careful here: The pendulum from hype to anti-hype can swing too far. It would be rash to dismiss this technology just because it has been oversold. The knee-jerk response when AI fails to live up to its hype is to say that progress has hit a wall. But that misunderstands how research and innovation in tech work. Progress has always moved in fits and starts. There are ways over, around, and under walls.

Take a step back from the GPT-5 launch. It came hot on the heels of a series of remarkable models that OpenAI had shipped in the previous months, including o1 and o3 (first-of-their-kind reasoning models that introduced the industry to a whole new paradigm) and Sora 2, which raised the bar for video generation once again. That doesn’t sound like hitting a wall to me.

AI is really good! Look at Nano Banana Pro, the new image generation model from Google DeepMind that can turn a book chapter into an infographic, and much more. It’s just there—for free—on your phone.

And yet you can’t help but wonder: When the wow factor is gone, what’s left? How will we view this technology a year or five from now? Will we think it was worth the colossal costs, both financial and environmental? 

With that in mind, here are four ways to think about the state of AI at the end of 2025: The start of a much-needed hype correction.

01: LLMs are not everything

In some ways, it is the hype around large language models, not AI as a whole, that needs correcting. It has become obvious that LLMs are not the doorway to artificial general intelligence, or AGI, a hypothetical technology that some insist will one day be able to do any (cognitive) task a human can.

Even an AGI evangelist like Ilya Sutskever, chief scientist and cofounder at the AI startup Safe Superintelligence and former chief scientist and cofounder at OpenAI, now highlights the limitations of LLMs, a technology he had a huge hand in creating. LLMs are very good at learning how to do a lot of specific tasks, but they do not seem to learn the principles behind those tasks, Sutskever said in an interview with Dwarkesh Patel in November.

It’s the difference between learning how to solve a thousand different algebra problems and learning how to solve any algebra problem. “The thing which I think is the most fundamental is that these models somehow just generalize dramatically worse than people,” Sutskever said.

It’s easy to imagine that LLMs can do anything because their use of language is so compelling. It is astonishing how well this technology can mimic the way people write and speak. And we are hardwired to see intelligence in things that behave in certain ways—whether it’s there or not. In other words, we have built machines with humanlike behavior and cannot resist seeing a humanlike mind behind them.

That’s understandable. LLMs have been part of mainstream life for only a few years. But in that time, marketers have preyed on our shaky sense of what the technology can really do, pumping up expectations and turbocharging the hype. As we live with this technology and come to understand it better, those expectations should fall back down to earth.  

02: AI is not a quick fix to all your problems

In July, researchers at MIT published a study that became a tentpole talking point in the disillusionment camp. The headline result was that a whopping 95% of businesses that had tried using AI had found zero value in it.  

The general thrust of that claim was echoed by other research, too. In November, a study by researchers at Upwork, a company that runs an online marketplace for freelancers, found that agents powered by top LLMs from OpenAI, Google DeepMind, and Anthropic failed to complete many straightforward workplace tasks by themselves.

This is miles off Altman’s prediction: “We believe that, in 2025, we may see the first AI agents ‘join the workforce’ and materially change the output of companies,” he wrote on his personal blog in January.

But what gets missed in that MIT study is that the researchers’ measure of success was pretty narrow. That 95% failure rate accounts for companies that had tried to implement bespoke AI systems but had not yet scaled them beyond the pilot stage after six months. It shouldn’t be too surprising that a lot of experiments with experimental technology don’t pan out straight away.

That number also does not include the use of LLMs by employees outside of official pilots. The MIT researchers found that around 90% of the companies they surveyed had a kind of AI shadow economy where workers were using personal chatbot accounts. But the value of that shadow economy was not measured.  

When the Upwork study looked at how well agents completed tasks together with people who knew what they were doing, success rates shot up. The takeaway seems to be that a lot of people are figuring out for themselves how AI might help them with their jobs.

That fits with something the AI researcher and influencer (and coiner of the term “vibe coding”) Andrej Karpathy has noted: Chatbots are better than the average human at a lot of different things (think of giving legal advice, fixing bugs, doing high school math), but they are not better than an expert human. Karpathy suggests this may be why chatbots have proved popular with individual consumers, helping non-experts with everyday questions and tasks, but they have not upended the economy, which would require outperforming skilled employees at their jobs.

That may change. For now, don’t be surprised that AI has not (yet) had the impact on jobs that boosters said it would. AI is not a quick fix, and it cannot replace humans. But there’s a lot to play for. The ways in which AI could be integrated into everyday workflows and business pipelines are still being tried out.   

03: Are we in a bubble? (If so, what kind of bubble?)

If AI is a bubble, is it like the subprime mortgage bubble of 2008 or the internet bubble of 2000? Because there’s a big difference.

The subprime bubble wiped out a big part of the economy, because when it burst it left nothing behind except debt and overvalued real estate. The dot-com bubble wiped out a lot of companies, which sent ripples across the world, but it left behind the infant internet—an international network of cables and a handful of startups, like Google and Amazon, that became the tech giants of today.  

Then again, maybe we’re in a bubble unlike either of those. After all, there’s no real business model for LLMs right now. We don’t yet know what the killer app will be, or if there will even be one. 

And many economists are concerned about the unprecedented amounts of money being sunk into the infrastructure required to build capacity and serve the projected demand. But what if that demand doesn’t materialize? Add to that the weird circularity of many of those deals—with Nvidia paying OpenAI to pay Nvidia, and so on—and it’s no surprise everybody’s got a different take on what’s coming. 

Some investors remain sanguine. In an interview with the Technology Business Programming Network podcast in November, Glenn Hutchins, cofounder of Silver Lake Partners, a major international private equity firm, gave a few reasons not to worry. “Every one of these data centers—almost all of them—has a solvent counterparty that is contracted to take all the output they’re built to suit,” he said. In other words, it’s not a case of “Build it and they’ll come”—the customers are already locked in. 

And, he pointed out, one of the biggest of those solvent counterparties is Microsoft. “Microsoft has the world’s best credit rating,” Hutchins said. “If you sign a deal with Microsoft to take the output from your data center, Satya is good for it.”

Many CEOs will be looking back at the dot-com bubble and trying to learn its lessons. Here’s one way to see it: The companies that went bust back then didn’t have the money to last the distance. Those that survived the crash thrived.

With that lesson in mind, AI companies today are trying to pay their way through what may or may not be a bubble. Stay in the race; don’t get left behind. Even so, it’s a desperate gamble.

But there’s another lesson too. Companies that might look like sideshows can turn into unicorns fast. Take Synthesia, which makes avatar generation tools for businesses. Nathan Benaich, cofounder of the VC firm Air Street Capital, admits that when he first heard about the company a few years ago, back when fear of deepfakes was rife, he wasn’t sure what its tech was for and thought there was no market for it.

“We didn’t know who would pay for lip-synching and voice cloning,” he says. “Turns out there’s a lot of people who wanted to pay for it.” Synthesia now has around 55,000 corporate customers and brings in around $150 million a year. In October, the company was valued at $4 billion.

04: ChatGPT was not the beginning, and it won’t be the end

ChatGPT was the culmination of a decade’s worth of progress in deep learning, the technology that underpins all of modern AI. The seeds of deep learning itself were planted in the 1980s. The field as a whole goes back at least to the 1950s. If progress is measured against that backdrop, generative AI has barely got going.

Meanwhile, research is at a fever pitch. There are more high-quality submissions to the world’s major AI conferences than ever before. This year, organizers of some of those conferences resorted to turning down papers that reviewers had already approved, just to manage numbers. (At the same time, preprint servers like arXiv have been flooded with AI-generated research slop.)

“It’s back to the age of research again,” Sutskever said in that Dwarkesh interview, talking about the current bottleneck with LLMs. That’s not a setback; that’s the start of something new.

“There’s always a lot of hype beasts,” says Benaich. But he thinks there’s an upside to that: Hype attracts the money and talent needed to make real progress. “You know, it was only like two or three years ago that the people who built these models were basically research nerds that just happened on something that kind of worked,” he says. “Now everybody who’s good at anything in technology is working on this.”

Where do we go from here?

The relentless hype hasn’t come just from companies drumming up business for their vastly expensive new technologies. There’s a large cohort of people—inside and outside the industry—who want to believe in the promise of machines that can read, write, and think. It’s a wild decades-old dream. 

But the hype was never sustainable—and that’s a good thing. We now have a chance to reset expectations and see this technology for what it really is—assess its true capabilities, understand its flaws, and take the time to learn how to apply it in valuable (and beneficial) ways. “We’re still trying to figure out how to invoke certain behaviors from this insanely high-dimensional black box of information and skills,” says Benaich.

This hype correction was long overdue. But know that AI isn’t going anywhere. We don’t even fully understand what we’ve built so far, let alone what’s coming next.

On April 28, 2022, at a highly anticipated concert in Spokane, Washington, the musician Paul McCartney astonished his audience with a groundbreaking application of AI: He began to perform with a lifelike depiction of his long-deceased musical partner, John Lennon. 

Using recent advances in audio and video processing, engineers had taken the pair’s final performance (London, 1969), separated Lennon’s voice and image from the original mix and restored them with lifelike clarity.

---

This story is part of MIT Technology Review’s Hype Correction package, a series that resets expectations about what AI is, what it makes possible, and where we go next.

---

For years, researchers like me had taught machines to “see” and “hear” in order to make such a moment possible. As McCartney and Lennon appeared to reunite across time and space, the arena fell silent; many in the crowd began to cry. As an AI scientist and lifelong Beatles fan, I felt profound gratitude that we could experience this truly life-changing moment. 

Later that year, the world was captivated by another major breakthrough: AI conversation. For the first time in history, systems capable of generating new, contextually relevant comments in real time, on virtually any subject, were widely accessible owing to the release of ChatGPT. Billions of people were suddenly able to interact with AI. This ignited the public’s imagination about what AI could be, bringing an explosion of creative ideas, hopes, and fears.

Having done my PhD on AI language generation (long considered niche), I was thrilled we had come this far. But the awe I felt was rivaled by my growing rage at the flood of media takes and self-appointed experts insisting that generative AI could do things it simply can’t, and warning that anyone who didn’t adopt it would be left behind.

This kind of hype has contributed to a frenzy of misunderstandings about what AI actually is and what it can and cannot do. Crucially, generative AI is a seductive distraction from the type of AI that is most likely to make your life better, or even save it: Predictive AI. In contrast to AI designed for generative tasks, predictive AI involves tasks with a finite, known set of answers; the system just has to process information to say which answer is right. A basic example is plant recognition: Point your phone camera at a plant and learn that it’s a Western sword fern. Generative tasks, in contrast, have no finite set of correct answers: The system must blend snippets of information it’s been trained on to create, for example, a novel picture of a fern. 

The generative AI technology involved in chatbots, face-swaps, and synthetic video makes for stunning demos, driving clicks and sales as viewers run wild with ideas that superhuman AI will be capable of bringing us abundance or extinction. Yet predictive AI has quietly been improving weather prediction and food safety, enabling higher-quality music production, helping to organize photos, and accurately predicting the fastest driving routes. We incorporate predictive AI into our everyday lives without evening thinking about it, a testament to its indispensable utility.

To get a sense of the immense progress on predictive AI and its future potential, we can look at the trajectory of the past 20 years. In 2005, we couldn’t get AI to tell the difference between a person and a pencil. By 2013, AI still couldn’t reliably detect a bird in a photo, and the difference between a pedestrian and a Coke bottle was massively confounding (this is how I learned that bottles do kind of look like people, if people had no heads). The thought of deploying these systems in the real world was the stuff of science fiction. 

Yet over the past 10 years, predictive AI has not only nailed bird detection down to the specific species; it has rapidly improved life-critical medical services like identifying problematic lesions and heart arrhythmia. Because of this technology, seismologists can predict earthquakes and meteorologists can predict flooding more reliably than ever before. Accuracy has skyrocketed for consumer-facing tech that detects and classifies everything from what song you’re thinking of when you hum a tune to which objects to avoid while you’re driving—making self-driving cars a reality. 

In the very near future, we should be able to accurately detect tumors and forecast hurricanes long before they can hurt anyone, realizing the lifelong hopes of people all over the world. That might not be as flashy as generating your own Studio Ghibli–ish film, but it’s definitely hype-worthy. 

Predictive AI systems have also been shown to be incredibly useful when they leverage certain generative techniques within a constrained set of options. Systems of this type are diverse, spanning everything from outfit visualization to cross-language translation. Soon, predictive-generative hybrid systems will make it possible to clone your own voice speaking another language in real time, an extraordinary aid for travel (with serious impersonation risks). There’s considerable room for growth here, but generative AI delivers real value when anchored by strong predictive methods.

To understand the difference between these two broad classes of AI, imagine yourself as an AI system tasked with showing someone what a cat looks like. You could adopt a generative approach, cutting and pasting small fragments from various cat images (potentially from sources that object) to construct a seemingly perfect depiction. The ability of modern generative AI to produce such a flawless collage is what makes it so astonishing.

Alternatively, you could take the predictive approach: Simply locate and point to an existing picture of a cat. That method is much less glamorous but more energy-efficient and more likely to be accurate, and it properly acknowledges the original source. Generative AI is designed to create things that look real; predictive AI identifies what is real. A misunderstanding that generative systems are retrieving things when they are actually creating them has led to grave consequences when text is involved, requiring the withdrawal of legal rulings and the retraction of scientific articles.

Driving this confusion is a tendency for people to hype AI without making it clear what kind of AI they’re talking about (I reckon many don’t know). It’s very easy to equate “AI” with generative AI, or even just language-generating AI, and assume that all other capabilities fall out from there. That fallacy makes a ton of sense: The term literally references “intelligence,” and our human understanding of what “intelligence” might be is often mediated by the use of language. (Spoiler: No one actually knows what intelligence is.) But the phrase “artificial intelligence” was intentionally designed in the 1950s to inspire awe and allude to something humanlike. Today, it just refers to a set of disparate technologies for processing digital data. Some of my friends find it helpful to call it “mathy maths” instead.

The bias toward treating generative AI as the most powerful and real form of AI is troubling given that it consumes considerably more energy than predictive AI systems. It also means using existing human work in AI products against the original creators’ wishes and replacing human jobs with AI systems whose capabilities their work made possible in the first place—without compensation. AI can be amazingly powerful, but that doesn’t mean creators should be ripped off. 

Watching this unfold as an AI developer within the tech industry, I’ve drawn important lessons for next steps. The widespread appeal of AI is clearly linked to the intuitive nature of conversation-based interactions. But this method of engagement currently overuses generative methods where predictive ones would suffice, resulting in an awkward situation that’s confusing for users while imposing heavy costs in energy consumption, exploitation, and job displacement. 

We have witnessed just a glimpse of AI’s full potential: The current excitement around AI reflects what it could be, not what it is. Generation-based approaches strain resources while still falling short on representation, accuracy, and the wishes of people whose work is folded into the system. 

If we can shift the spotlight from the hype around generative technologies to the predictive advances already transforming daily life, we can build AI that is genuinely useful, equitable, and sustainable. The systems that help doctors catch diseases earlier, help scientists forecast disasters sooner, and help everyday people navigate their lives more safely are the ones poised to deliver the greatest impact. 

The future of beneficial AI will not be defined by the flashiest demos but by the quiet, rigorous progress that makes technology trustworthy. And if we build on that foundation—pairing predictive strength with more mature data practices and intuitive natural-language interfaces—AI can finally start living up to the promise that many people perceive today.

Dr. Margaret Mitchell is a computer science researcher and chief ethics scientist at AI startup Hugging Face. She has worked in the technology industry for 15 years, and has published over 100 papers on natural language generation, assistive technology, computer vision, and AI ethics. Her work has received numerous awards and has been implemented by multiple technology companies.

When the generative AI boom took off in 2022, Rudi Miller and her law school classmates were suddenly gripped with anxiety. “Before graduating, there was discussion about what the job market would look like for us if AI became adopted,” she recalls. 

So when it came time to choose a speciality, Miller—now a junior associate at the law firm Orrick—decided to become a litigator, the kind of lawyer who represents clients in court. She hoped the courtroom would be the last human stage. “Judges haven’t allowed ChatGPT-enabled robots to argue in court yet,” she says.

---

This story is part of MIT Technology Review’s Hype Correction package, a series that resets expectations about what AI is, what it makes possible, and where we go next.

---

She had reason to be worried. The artificial-intelligence job apocalypse seemed to be coming for lawyers. In March 2023, researchers reported that GPT-4 had smashed the Uniform Bar Exam. That same month, an industry report predicted that 44% of legal work could be automated. The legal tech industry entered a boom as law firms began adopting generative AI to mine mountains of documents and draft contracts, work ordinarily done by junior associates. Last month, the law firm Clifford Chance axed 10% of its staff in London, citing increased use of AI as a reason.

But for all the hype, LLMs are still far from thinking like lawyers—let alone replacing them. The models continue to hallucinate case citations, struggle to navigate gray areas of the law and reason about novel questions, and stumble when they attempt to synthesize information scattered across statutes, regulations, and court cases. And there are deeper institutional reasons to think the models could struggle to supplant legal jobs. While AI is reshaping the grunt work of the profession, the end of lawyers may not be arriving anytime soon.

The big experiment

The legal industry has long been defined by long hours and grueling workloads, so the promise of superhuman efficiency is appealing. Law firms are experimenting with general-purpose tools like ChatGPT and Microsoft Copilot and specialized legal tools like Harvey and Thomson Reuters’ CoCounsel, with some building their own in-house tools on top of frontier models. They’re rolling out AI boot camps and letting associates bill hundreds of hours to AI experimentation. As of 2024, 47.8% of attorneys at law firms employing 500 or more lawyers used AI, according to the American Bar Association. 

But lawyers say that LLMs are a long way from reasoning well enough to replace them. Lucas Hale, a junior associate at McDermott Will & Schulte, has been embracing AI for many routine chores. He uses Relativity to sift through long documents and Microsoft Copilot for drafting legal citations. But when he turns to ChatGPT with a complex legal question, he finds the chatbot spewing hallucinations, rambling off topic, or drawing a blank.

“In the case where we have a very narrow question or a question of first impression for the court,” he says, referring to a novel legal question that a court has never decided before, “that’s the kind of thinking that the tool can’t do.”

Much of Lucas’s work involves creatively applying the law to new fact patterns. “Right now, I don’t think very much of the work that litigators do, at least not the work that I do, can be outsourced to an AI utility,” he says.

Allison Douglis, a senior associate at Jenner & Block, uses an LLM to kick off her legal research. But the tools only take her so far. “When it comes to actually fleshing out and developing an argument as a litigator, I don’t think they’re there,” she says. She has watched the models hallucinate case citations and fumble through ambiguous areas of the law.

“Right now, I would much rather work with a junior associate than an AI tool,” she says. “Unless they get extraordinarily good very quickly, I can’t imagine that changing in the near future.”

Beyond the bar

The legal industry has seemed ripe for an AI takeover ever since ChatGPT’s triumph on the bar exam. But passing a standardized test isn’t the same as practicing law. The exam tests whether people can memorize legal rules and apply them to hypothetical situations—not whether they can exercise strategic judgment in complicated realities or craft arguments in uncharted legal territory. And models can be trained to ace benchmarks without genuinely improving their reasoning.

But new benchmarks are aiming to better measure the models’ ability to do legal work in the real world. The Professional Reasoning Benchmark, published by ScaleAI in November, evaluated leading LLMs on legal and financial tasks designed by professionals in the field. The study found that the models have critical gaps in their reliability for professional adoption, with the best-performing model scoring only 37% on the most difficult legal problems, meaning it met just over a third of possible points on the evaluation criteria. The models frequently made inaccurate legal judgments, and if they did reach correct conclusions, they did so through incomplete or opaque reasoning processes. 

“The tools actually are not there to basically substitute [for] your lawyer,” says Afra Feyza Akyurek, the lead author of the paper. “Even though a lot of people think that LLMs have a good grasp of the law, it’s still lagging behind.” 

The paper builds on other benchmarks measuring the models’ performance on economically valuable work. The AI Productivity Index, published by the data firm Mercor in September and updated in December, found that the models have “substantial limitations” in performing legal work. The best-performing model scored 77.9% on legal tasks, meaning it satisfied roughly four out of five evaluation criteria. A model with such a score might generate substantial economic value in some industries, but in fields where errors are costly, it may not be useful at all, the early version of the study noted.  

Professional benchmarks are a big step forward in evaluating the LLMs’ real-world capabilities, but they may still not capture what lawyers actually do. “These questions, although more challenging than those in past benchmarks, still don’t fully reflect the kinds of subjective, extremely challenging questions lawyers tackle in real life,” says Jon Choi, a law professor at the University of Washington School of Law, who coauthored a study on legal benchmarks in 2023. 

Unlike math or coding, in which LLMs have made significant progress, legal reasoning may be challenging for the models to learn. The law deals with messy real-world problems, riddled with ambiguity and subjectivity, that often have no right answer, says Choi. Making matters worse, a lot of legal work isn’t recorded in ways that can be used to train the models, he says. When it is, documents can span hundreds of pages, scattered across statutes, regulations, and court cases that exist in a complex hierarchy.  

But a more fundamental limitation might be that LLMs are simply not trained to think like lawyers. “The reasoning models still don’t fully reason about problems like we humans do,” says Julian Nyarko, a law professor at Stanford Law School. The models may lack a mental model of the world—the ability to simulate a scenario and predict what will happen—and that capability could be at the heart of complex legal reasoning, he says. It’s possible that the current paradigm of LLMs trained on next-word prediction gets us only so far.  

The jobs remain

Despite early signs that AI is beginning to affect entry-level workers, labor statistics have yet to show that lawyers are being displaced. 93.4% of law school graduates in 2024 were employed within 10 months of graduation—the highest rate on record—according to the National Association for Law Placement. The number of graduates working in law firms rose by 13% from 2023 to 2024. 

For now, law firms are slow to shrink their ranks. “We’re not reducing headcounts at this point,” said Amy Ross, the chief of attorney talent at the law firm Ropes & Gray. 

Even looking ahead, the effects could be incremental. “I will expect some impact on the legal profession’s labor market, but not major,” says Mert Demirer, an economist at MIT. “AI is going to be very useful in terms of information discovery and summary,” he says, but for complex legal tasks, “the law’s low risk tolerance, plus the current capabilities of AI, are going to make that case less automatable at this point.” Capabilities may evolve over time, but that’s a big unknown.

It’s not just that the models themselves are not ready to replace junior lawyers. Institutional barriers may also shape how AI is deployed. Higher productivity reduces billable hours, challenging the dominant business model of law firms. Liability looms large for lawyers, and clients may still want a human on the hook. Regulations could also constrain how lawyers use the technology.

Still, as AI takes on some associate work, law firms may need to reinvent their training system. “When junior work dries up, you have to have a more formal way of teaching than hoping that an apprenticeship works,” says Ethan Mollick, a management professor at the Wharton School of the University of Pennsylvania.

Zach Couger, a junior associate at McDermott Will & Schulte, leans on ChatGPT to comb through piles of contracts he once slogged through by hand. He can’t imagine going back to doing the job himself, but he wonders what he’s missing. 

“I’m worried that I’m not getting the same reps that senior attorneys got,” he says, referring to the repetitive training that has long defined the early experiences of lawyers. “On the other hand, it is very nice to have a semi–knowledge expert to just ask questions to that’s not a partner who’s also very busy.” 

Even though an AI job apocalypse looks distant, the uncertainty sticks with him. Lately, Couger finds himself staying up late, wondering if he could be part of the last class of associates at big law firms: “I may be the last plane out.”

MIT Technology Review Explains: Let our writers untangle the complex, messy world of technology to help you understand what’s coming next. You can read more from the series here.

In July, a widely cited MIT study claimed that 95% of organizations that invested in generative AI were getting “zero return.” Tech stocks briefly plunged. While the study itself was more nuanced than the headlines, for many it still felt like the first hard data point confirming what skeptics had muttered for months: Hype around AI might be outpacing reality.

Then, in August, OpenAI CEO Sam Altman said what everyone in Silicon Valley had been whispering. “Are we in a phase where investors as a whole are overexcited about AI?” he said during a press dinner I attended. “My opinion is yes.” 

---

This story is part of MIT Technology Review’s Hype Correction package, a series that resets expectations about what AI is, what it makes possible, and where we go next.

---

He compared the current moment to the dot-com bubble. “When bubbles happen, smart people get overexcited about a kernel of truth,” he explained. “Tech was really important. The internet was a really big deal. People got overexcited.” 

With those comments, it was off to the races. The next day’s stock market dip was attributed to the sentiment he shared. The question “Are we in an AI bubble?” became inescapable.

Who thinks it is a bubble? 

The short answer: Lots of people. But not everyone agrees on who or what is overinflated. Tech leaders are using this moment of fear to take shots at their rivals and position themselves as clear winners on the other side. How they describe the bubble depends on where their company sits.

When I asked Meta CEO Mark Zuckerberg about the AI bubble in September, he ran through the historical analogies of past bubbles—railroads, fiber for the internet, the dot-com boom—and noted that in each case, “the infrastructure gets built out, people take on too much debt, and then you hit some blip … and then a lot of the companies end up going out of business.”

But Zuckerberg’s prescription wasn’t for Meta to pump the brakes. It was to keep spending: “If we end up misspending a couple of hundred billion dollars, I think that that is going to be very unfortunate, obviously. But I’d say the risk is higher on the other side.”

Bret Taylor, the chairman of OpenAI and CEO of the AI startup Sierra, uses a mental model from the late ’90s to help navigate this AI bubble. “I think the closest analogue to this AI wave is the dot-com boom or bubble, depending on your level of pessimism,” he recently told me. Back then, he explained, everyone knew e-commerce was going to be big, but there was a massive difference between Buy.com and Amazon. Taylor and others have been trying to position themselves as today’s Amazon.

Still others are arguing that the pain will be widespread. Google CEO Sundar Pichai told the BBC this month that there’s “some irrationality” in the current boom. Asked whether Google would be immune to a bubble bursting, he warned, “I think no company is going to be immune, including us.”

What’s inflating the bubble?

Companies are raising enormous sums of money and seeing unprecedented valuations. Much of that money, in turn, is going toward the buildout of massive data centers—on which both private companies like OpenAI and Elon Musk’s xAI and public ones such as Meta and Google are spending heavily. OpenAI has pledged that it will spend $500 billion to build AI data centers, more than 15 times what was spent on the Manhattan Project.

This eye-popping spending on AI data centers isn’t entirely detached from reality. The leaders of the top AI companies all stress that they’re bottlenecked by their limited access to computing power. You hear it constantly when you talk to them. Startups can’t get the GPU allocations they need. Hyperscalers are rationing compute, saving it for their best customers.

If today’s AI market is as brutally supply-constrained as tech leaders claim, perhaps aggressive infrastructure buildouts are warranted. But some of the numbers are too large to comprehend. Sam Altman has told employees that OpenAI’s moonshot goal is to build 250 gigawatts of computing capacity by 2033, roughly equaling India’s total national electricity demand. Such a plan would cost more than $12 trillion by today’s standards.

“I do think there’s real execution risk,” OpenAI president and cofounder Greg Brockman recently told me about the company’s aggressive infrastructure goals. “Everything we say about the future, we see that it’s a possibility. It is not a certainty, but I don’t think the uncertainty comes from scientific questions. It’s a lot of hard work.”

Who is exposed, and who is to blame?

It depends on who you ask. During the August press dinner, where he made his market-moving comments, Altman was blunt about where he sees the excess. He said it’s “insane” that some AI startups with “three people and an idea” are receiving funding at such high valuations. “That’s not rational behavior,” he said. “Someone’s gonna get burned there, I think.” As Safe Superintelligence cofounder (and former OpenAI chief scientist and cofounder) Ilya Sutskever put it on a recent podcast: Silicon Valley has “more companies than ideas.”

Demis Hassabis, the CEO of Google DeepMind, offered a similar diagnosis when I spoke with him in November. “It feels like there’s obviously a bubble in the private market,” he said. “You look at seed rounds with just nothing being tens of billions of dollars. That seems a little unsustainable.”

Anthropic CEO Dario Amodei also struck at his competition during the New York Times DealBook Summit in early December. He said he feels confident about the technology itself but worries about how others are behaving on the business side: “On the economic side, I have my concerns where, even if the technology fulfills all its promises, I think there are players in the ecosystem who, if they just make a timing error, they just get it off by a little bit, bad things could happen.”

He stopped short of naming Sam Altman and OpenAI, but the implication was clear. “There are some players who are YOLOing,” he said. “Let’s say you’re a person who just kind of constitutionally wants to YOLO things or just likes big numbers. Then you may turn the dial too far.”

Amodei also flagged “circular deals,” or the increasingly common arrangements where chip suppliers like Nvidia invest in AI companies that then turn around and spend those funds on their chips. Anthropic has done some of these, he said, though “not at the same scale as some other players.” (OpenAI is at the center of a number of such deals, as are Nvidia, CoreWeave, and a roster of other players.) 

The danger, he explained, comes when the numbers get too big: “If you start stacking these where they get to huge amounts of money, and you’re saying, ’By 2027 or 2028 I need to make $200 billion a year,’ then yeah, you can overextend yourself.”

Zuckerberg shared a similar message at an internal employee Q&A session after Meta’s last earnings call. He noted that unprofitable startups like OpenAI and Anthropic risk bankruptcy if they misjudge the timing of their investments, but Meta has the advantage of strong cash flow, he reassured staff.

How could a bubble burst?

My conversations with tech executives and investors suggest that the bubble will be most likely to pop if overfunded startups can’t turn a profit or grow into their lofty valuations. This bubble could last longer than than past ones, given that private markets aren’t traded on public markets and therefore move more slowly, but the ripple effects will still be profound when the end comes. 

If companies making grand commitments to data center buildouts no longer have the revenue growth to support them, the headline deals that have propped up the stock market come into question. Anthropic’s Amodei illustrated the problem during his DealBook Summit appearance, where he said the multi-year data center commitments he has to make combine with the company’s rapid, unpredictable revenue growth rate to create a “cone of uncertainty” about how much to spend.

The two most prominent private players in AI, OpenAI and Anthropic, have yet to turn a profit. A recent Deutsche Bank chart put the situation in stark historical context. Amazon burned through $3 billion before becoming profitable. Tesla, around $4 billion. Uber, $30 billion. OpenAI is projected to burn through $140 billion by 2029, while Anthropic is expected to burn $20 billion by 2027.

Consultants at Bain estimate that the wave of AI infrastructure spending will require $2 trillion in annual AI revenue by 2030 just to justify the investment. That’s more than the combined 2024 revenue of Amazon, Apple, Alphabet, Microsoft, Meta, and Nvidia. When I talk to leaders of these large tech companies, they all agree that their sprawling businesses can absorb an expensive miscalculation about the returns from their AI infrastructure buildouts. It’s all the other companies that are either highly leveraged with debt or just unprofitable—even OpenAI and Anthropic—that they worry about. 

Still, given the level of spending on AI, it still needs a viable business model beyond subscriptions, which won’t be able to  drive profits from billions of people’s eyeballs like the ad-driven businesses that have defined the last 20 years of the internet. Even the largest tech companies know they need to ship the world-changing agents they keep hyping: AI that can fully replace coworkers and complete tasks in the real world.

For now, investors are mostly buying into the hype of the powerful AI systems that these data center buildouts will supposedly unlock in the future. At some point the biggest spenders, like OpenAI, will need to show investors that the money spent on the infrastructure buildout was worth it.

There’s also still a lot of uncertainty about the technical direction that AI is heading in. LLMs are expected to remain critical to more advanced AI systems, but industry leaders can’t seem to agree on which additional breakthroughs are needed to achieve artificial general intelligence, or AGI. Some are betting on new kinds of AI that can understand the physical world, while others are focused on training AI to learn in a general way, like a human. In other words, what if all this unprecedented spending turns out to have been backing the wrong horse?

The question now

What makes this moment surreal is the honesty. The same people pouring billions into AI will openly tell you it might all come crashing down. 

Taylor framed it as two truths existing at once. “I think it is both true that AI will transform the economy,” he told me, “and I think we’re also in a bubble, and a lot of people will lose a lot of money. I think both are absolutely true at the same time.”

He compared it to the internet. Webvan failed, but Instacart succeeded years later with essentially the same idea. If you were an Amazon shareholder from its IPO to now, you’re looking pretty good. If you were a Webvan shareholder, you probably feel differently. 

“When the dust settles and you see who the winners are, society benefits from those inventions,” Amazon founder Jeff Bezos said in October. “This is real. The benefit to society from AI is going to be gigantic.”

Goldman Sachs says the AI boom now looks the way tech stocks did in 1997, several years before the dot-com bubble actually burst. The bank flagged five warning signs seen in the late 1990s that investors should watch now: peak investment spending, falling corporate profits, rising corporate debt, Fed rate cuts, and widening credit spreads. We’re probably not at 1999 levels yet. But the imbalances are building fast. Michael Burry, who famously called the 2008 housing bubble collapse (as seen in the film The Big Short), recently compared the AI boom to the 1990s dot-com bubble too.

Maybe AI will save us from our own irrational exuberance. But for now, we’re living in an in-between moment when everyone knows what’s coming but keeps blowing more air into the balloon anyway. As Altman put it that night at dinner: “Someone is going to lose a phenomenal amount of money. We don’t know who.”

Alex Heath is the author of Sources, a newsletter about the AI race, and the cohost of ACCESS, a podcast about the tech industry’s inside conversations. Previously, he was deputy editor at The Verge.

The microwave-size instrument at Lila Sciences in Cambridge, Massachusetts, doesn’t look all that different from others that I’ve seen in state-of-the-art materials labs. Inside its vacuum chamber, the machine zaps a palette of different elements to create vaporized particles, which then fly through the chamber and land to create a thin film, using a technique called sputtering. What sets this instrument apart is that artificial intelligence is running the experiment; an AI agent, trained on vast amounts of scientific literature and data, has determined the recipe and is varying the combination of elements. 

Later, a person will walk the samples, each containing multiple potential catalysts, over to a different part of the lab for testing. Another AI agent will scan and interpret the data, using it to suggest another round of experiments to try to optimize the materials’ performance.  

---

This story is part of MIT Technology Review’s Hype Correction package, a series that resets expectations about what AI is, what it makes possible, and where we go next.

---

For now, a human scientist keeps a close eye on the experiments and will approve the next steps on the basis of the AI’s suggestions and the test results. But the startup is convinced this AI-controlled machine is a peek into the future of materials discovery—one in which autonomous labs could make it far cheaper and faster to come up with novel and useful compounds. 

Flush with hundreds of millions of dollars in new funding, Lila Sciences is one of AI’s latest unicorns. The company is on a larger mission to use AI-run autonomous labs for scientific discovery—the goal is to achieve what it calls scientific superintelligence. But I’m here this morning to learn specifically about the discovery of new materials. 

We desperately need better materials to solve our problems. We’ll need improved electrodes and other parts for more powerful batteries; compounds to more cheaply suck carbon dioxide out of the air; and better catalysts to make green hydrogen and other clean fuels and chemicals. And we will likely need novel materials like higher-temperature superconductors, improved magnets, and different types of semiconductors for a next generation of breakthroughs in everything from quantum computing to fusion power to AI hardware. 

But materials science has not had many commercial wins in the last few decades. In part because of its complexity and the lack of successes, the field has become something of an innovation backwater, overshadowed by the more glamorous—and lucrative—search for new drugs and insights into biology.

The idea of using AI for materials discovery is not exactly new, but it got a huge boost in 2020 when DeepMind showed that its AlphaFold2 model could accurately predict the three-dimensional structure of proteins. Then, in 2022, came the success and popularity of ChatGPT. The hope that similar AI models using deep learning could aid in doing science captivated tech insiders. Why not use our new generative AI capabilities to search the vast chemical landscape and help simulate atomic structures, pointing the way to new substances with amazing properties?

“Simulations can be super powerful for framing problems and understanding what is worth testing in the lab. But there’s zero problems we can ever solve in the real world with simulation alone.”

John Gregoire, Lila Sciences, chief autonomous science officer

Researchers touted an AI model that had reportedly discovered “millions of new materials.” The money began pouring in, funding a host of startups. But so far there has been no “eureka” moment, no ChatGPT-like breakthrough—no discovery of new miracle materials or even slightly better ones.

The startups that want to find useful new compounds face a common bottleneck: By far the most time-consuming and expensive step in materials discovery is not imagining new structures but making them in the real world. Before trying to synthesize a material, you don’t know if, in fact, it can be made and is stable, and many of its properties remain unknown until you test it in the lab.

“Simulations can be super powerful for kind of framing problems and understanding what is worth testing in the lab,” says John Gregoire, Lila Sciences’ chief autonomous science officer. “But there’s zero problems we can ever solve in the real world with simulation alone.” 

Startups like Lila Sciences have staked their strategies on using AI to transform experimentation and are building labs that use agents to plan, run, and interpret the results of experiments to synthesize new materials. Automation in laboratories already exists. But the idea is to have AI agents take it to the next level by directing autonomous labs, where their tasks could include designing experiments and controlling the robotics used to shuffle samples around. And, most important, companies want to use AI to vacuum up and analyze the vast amount of data produced by such experiments in the search for clues to better materials.

If they succeed, these companies could shorten the discovery process from decades to a few years or less, helping uncover new materials and optimize existing ones. But it’s a gamble. Even though AI is already taking over many laboratory chores and tasks, finding new—and useful—materials on its own is another matter entirely. 

Innovation backwater

I have been reporting about materials discovery for nearly 40 years, and to be honest, there have been only a few memorable commercial breakthroughs, such as lithium-­ion batteries, over that time. There have been plenty of scientific advances to write about, from perovskite solar cells to graphene transistors to metal-­organic frameworks (MOFs), materials based on an intriguing type of molecular architecture that recently won its inventors a Nobel Prize. But few of those advances—including MOFs—have made it far out of the lab. Others, like quantum dots, have found some commercial uses, but in general, the kinds of life-changing inventions created in earlier decades have been lacking. 

Blame the amount of time (typically 20 years or more) and the hundreds of millions of dollars it takes to make, test, optimize, and manufacture a new material—and the industry’s lack of interest in spending that kind of time and money in low-margin commodity markets. Or maybe we’ve just run out of ideas for making stuff.

The need to both speed up that process and find new ideas is the reason researchers have turned to AI. For decades, scientists have used computers to design potential materials, calculating where to place atoms to form structures that are stable and have predictable characteristics. It’s worked—but only kind of. Advances in AI have made that computational modeling far faster and have promised the ability to quickly explore a vast number of possible structures. Google DeepMind, Meta, and Microsoft have all launched efforts to bring AI tools to the problem of designing new materials. 

But the limitations that have always plagued computational modeling of new materials remain. With many types of materials, such as crystals, useful characteristics often can’t be predicted solely by calculating atomic structures.

To uncover and optimize those properties, you need to make something real. Or as Rafael Gómez-Bombarelli, one of Lila’s cofounders and an MIT professor of materials science, puts it: “Structure helps us think about the problem, but it’s neither necessary nor sufficient for real materials problems.”

Perhaps no advance exemplified the gap between the virtual and physical worlds more than DeepMind’s announcement in late 2023 that it had used deep learning to discover “millions of new materials,” including 380,000 crystals that it declared “the most stable, making them promising candidates for experimental synthesis.” In technical terms, the arrangement of atoms represented a minimum energy state where they were content to stay put. This was “an order-of-magnitude expansion in stable materials known to humanity,” the DeepMind researchers proclaimed.

To the AI community, it appeared to be the breakthrough everyone had been waiting for. The DeepMind research not only offered a gold mine of possible new materials, it also created powerful new computational methods for predicting a large number of structures.

But some materials scientists had a far different reaction. After closer scrutiny, researchers at the University of California, Santa Barbara, said they’d found “scant evidence for compounds that fulfill the trifecta of novelty, credibility, and utility.” In fact, the scientists reported, they didn’t find any truly novel compounds among the ones they looked at; some were merely “trivial” variations of known ones. The scientists appeared particularly peeved that the potential compounds were labeled materials. They wrote: “We would respectfully suggest that the work does not report any new materials but reports a list of proposed compounds. In our view, a compound can be called a material when it exhibits some functionality and, therefore, has potential utility.”

Some of the imagined crystals simply defied the conditions of the real world. To do computations on so many possible structures, DeepMind researchers simulated them at absolute zero, where atoms are well ordered; they vibrate a bit but don’t move around. At higher temperatures—the kind that would exist in the lab or anywhere in the world—the atoms fly about in complex ways, often creating more disorderly crystal structures. A number of the so-called novel materials predicted by DeepMind appeared to be well-ordered versions of disordered ones that were already known. 

More generally, the DeepMind paper was simply another reminder of how challenging it is to capture physical realities in virtual simulations—at least for now. Because of the limitations of computational power, researchers typically perform calculations on relatively few atoms. Yet many desirable properties are determined by the microstructure of the materials—at a scale much larger than the atomic world. And some effects, like high-temperature superconductivity or even the catalysis that is key to many common industrial processes, are far too complex or poorly understood to be explained by atomic simulations alone.

A common language

Even so, there are signs that the divide between simulations and experimental work is beginning to narrow. DeepMind, for one, says that since the release of the 2023 paper it has been working with scientists in labs around the world to synthesize AI-identified compounds and has achieved some success. Meanwhile, a number of the startups entering the space are looking to combine computational and experimental expertise in one organization. 

One such startup is Periodic Labs, cofounded by Ekin Dogus Cubuk, a physicist who led the scientific team that generated the 2023 DeepMind headlines, and by Liam Fedus, a co-creator of ChatGPT at OpenAI. Despite its founders’ background in computational modeling and AI software, the company is building much of its materials discovery strategy around synthesis done in automated labs. 

The vision behind the startup is to link these different fields of expertise by using large language models that are trained on scientific literature and able to learn from ongoing experiments. An LLM might suggest the recipe and conditions to make a compound; it can also interpret test data and feed additional suggestions to the startup’s chemists and physicists. In this strategy, simulations might suggest possible material candidates, but they are also used to help explain the experimental results and suggest possible structural tweaks.

The grand prize would be a room-temperature superconductor, a material that could transform computing and electricity but that has eluded scientists for decades.

Periodic Labs, like Lila Sciences, has ambitions beyond designing and making new materials. It wants to “create an AI scientist”—specifically, one adept at the physical sciences. “LLMs have gotten quite good at distilling chemistry information, physics information,” says Cubuk, “and now we’re trying to make it more advanced by teaching it how to do science—for example, doing simulations, doing experiments, doing theoretical modeling.”

The approach, like that of Lila Sciences, is based on the expectation that a better understanding of the science behind materials and their synthesis will lead to clues that could help researchers find a broad range of new ones. One target for Periodic Labs is materials whose properties are defined by quantum effects, such as new types of magnets. The grand prize would be a room-temperature superconductor, a material that could transform computing and electricity but that has eluded scientists for decades.

Superconductors are materials in which electricity flows without any resistance and, thus, without producing heat. So far, the best of these materials become superconducting only at relatively low temperatures and require significant cooling. If they can be made to work at or close to room temperature, they could lead to far more efficient power grids, new types of quantum computers, and even more practical high-speed magnetic-levitation trains. 

The failure to find a room-­temperature superconductor is one of the great disappointments in materials science over the last few decades. I was there when President Reagan spoke about the technology in 1987, during the peak hype over newly made ceramics that became superconducting at the relatively balmy temperature of 93 Kelvin (that’s −292 °F), enthusing that they “bring us to the threshold of a new age.” There was a sense of optimism among the scientists and businesspeople in that packed ballroom at the Washington Hilton as Reagan anticipated “a host of benefits, not least among them a reduced dependence on foreign oil, a cleaner environment, and a stronger national economy.” In retrospect, it might have been one of the last times that we pinned our economic and technical aspirations on a breakthrough in materials.

The promised new age never came. Scientists still have not found a material that becomes superconducting at room temperatures, or anywhere close, under normal conditions. The best existing superconductors are brittle and tend to make lousy wires.

One of the reasons that finding higher-­temperature superconductors has been so difficult is that no theory explains the effect at relatively high temperatures—or can predict it simply from the placement of atoms in the structure. It will ultimately fall to lab scientists to synthesize any interesting candidates, test them, and search the resulting data for clues to understanding the still puzzling phenomenon. Doing so, says Cubuk, is one of the top priorities of Periodic Labs. 

AI in charge

It can take a researcher a year or more to make a crystal structure for the first time. Then there are typically years of further work to test its properties and figure out how to make the larger quantities needed for a commercial product. 

Startups like Lila Sciences and Periodic Labs are pinning their hopes largely on the prospect that AI-directed experiments can slash those times. One reason for the optimism is that many labs have already incorporated a lot of automation, for everything from preparing samples to shuttling test items around. Researchers routinely use robotic arms, software, automated versions of microscopes and other analytical instruments, and mechanized tools for manipulating lab equipment.

The automation allows, among other things, for high-throughput synthesis, in which multiple samples with various combinations of ingredients are rapidly created and screened in large batches, greatly speeding up the experiments.

The idea is that using AI to plan and run such automated synthesis can make it far more systematic and efficient. AI agents, which can collect and analyze far more data than any human possibly could, can use real-time information to vary the ingredients and synthesis conditions until they get a sample with the optimal properties. Such AI-directed labs could do far more experiments than a person and could be far smarter than existing systems for high-throughput synthesis. 

But so-called self-driving labs for materials are still a work in progress.

Many types of materials require solid-­state synthesis, a set of processes that are far more difficult to automate than the liquid-­handling activities that are commonplace in making drugs. You need to prepare and mix powders of multiple inorganic ingredients in the right combination for making, say, a catalyst and then decide how to process the sample to create the desired structure—for example, identifying the right temperature and pressure at which to carry out the synthesis. Even determining what you’ve made can be tricky.

In 2023, the A-Lab at Lawrence Berkeley National Laboratory claimed to be the first fully automated lab to use inorganic powders as starting ingredients. Subsequently, scientists reported that the autonomous lab had used robotics and AI to synthesize and test 41 novel materials, including some predicted in the DeepMind database. Some critics questioned the novelty of what was produced and complained that the automated analysis of the materials was not up to experimental standards, but the Berkeley researchers defended the effort as simply a demonstration of the autonomous system’s potential.

“How it works today and how we envision it are still somewhat different. There’s just a lot of tool building that needs to be done,” says Gerbrand Ceder, the principal scientist behind the A-Lab. 

AI agents are already getting good at doing many laboratory chores, from preparing recipes to interpreting some kinds of test data—finding, for example, patterns in a micrograph that might be hidden to the human eye. But Ceder is hoping the technology could soon “capture human decision-making,” analyzing ongoing experiments to make strategic choices on what to do next. For example, his group is working on an improved synthesis agent that would better incorporate what he calls scientists’ “diffused” knowledge—the kind gained from extensive training and experience. “I imagine a world where people build agents around their expertise, and then there’s sort of an uber-model that puts it together,” he says. “The uber-model essentially needs to know what agents it can call on and what they know, or what their expertise is.”

“In one field that I work in, solid-state batteries, there are 50 papers published every day. And that is just one field that I work in. The A I revolution is about finally gathering all the scientific data we have.”

Gerbrand Ceder, principal scientist, A-Lab

One of the strengths of AI agents is their ability to devour vast amounts of scientific literature. “In one field that I work in, solid-­state batteries, there are 50 papers published every day. And that is just one field that I work in,” says Ceder. It’s impossible for anyone to keep up. “The AI revolution is about finally gathering all the scientific data we have,” he says. 

Last summer, Ceder became the chief science officer at an AI materials discovery startup called Radical AI and took a sabbatical from the University of California, Berkeley, to help set up its self-driving labs in New York City. A slide deck shows the portfolio of different AI agents and generative models meant to help realize Ceder’s vision. If you look closely, you can spot an LLM called the “orchestrator”—it’s what CEO Joseph Krause calls the “head honcho.” 

New hope

So far, despite the hype around the use of AI to discover new materials and the growing momentum—and money—behind the field, there still has not been a convincing big win. There is no example like the 2016 victory of DeepMind’s AlphaGo over a Go world champion. Or like AlphaFold’s achievement in mastering one of biomedicine’s hardest and most time-consuming chores, predicting 3D structures of proteins. 

The field of materials discovery is still waiting for its moment. It could come if AI agents can dramatically speed the design or synthesis of practical materials, similar to but better than what we have today. Or maybe the moment will be the discovery of a truly novel one, such as a room-­temperature superconductor.

With or without such a breakthrough moment, startups face the challenge of trying to turn their scientific achievements into useful materials. The task is particularly difficult because any new materials would likely have to be commercialized in an industry dominated by large incumbents that are not particularly prone to risk-taking.

Susan Schofer, a tech investor and partner at the venture capital firm SOSV, is cautiously optimistic about the field. But Schofer, who spent several years in the mid-2000s as a catalyst researcher at one of the first startups using automation and high-throughput screening for materials discovery (it didn’t survive), wants to see some evidence that the technology can translate into commercial successes when she evaluates startups to invest in.  

In particular, she wants to see evidence that the AI startups are already “finding something new, that’s different, and know how they are going to iterate from there.” And she wants to see a business model that captures the value of new materials. She says, “I think the ideal would be: I got a spec from the industry. I know what their problem is. We’ve defined it. Now we’re going to go build it. Now we have a new material that we can sell, that we have scaled up enough that we’ve proven it. And then we partner somehow to manufacture it, but we get revenue off selling the material.”

Schofer says that while she gets the vision of trying to redefine science, she’d advise startups to “show us how you’re going to get there.” She adds, “Let’s see the first steps.”

Demonstrating those first steps could be essential in enticing large existing materials companies to embrace AI technologies more fully. Corporate researchers in the industry have been burned before—by the promise over the decades that increasingly powerful computers will magically design new materials; by combinatorial chemistry, a fad that raced through materials R&D labs in the early 2000s with little tangible result; and by the promise that synthetic biology would make our next generation of chemicals and materials.

More recently, the materials community has been blanketed by a new hype cycle around AI. Some of that hype was fueled by the 2023 DeepMind announcement of the discovery of “millions of new materials,” a claim that, in retrospect, clearly overpromised. And it was further fueled when an MIT economics student posted a paper in late 2024 claiming that a large, unnamed corporate R&D lab had used AI to efficiently invent a slew of new materials. AI, it seemed, was already revolutionizing the industry.

A few months later, the MIT economics department concluded that “the paper should be withdrawn from public discourse.” Two prominent MIT economists who are acknowledged in a footnote in the paper added that they had “no confidence in the provenance, reliability or validity of the data and the veracity of the research.”

Can AI move beyond the hype and false hopes and truly transform materials discovery? Maybe. There is ample evidence that it’s changing how materials scientists work, providing them—if nothing else—with useful lab tools. Researchers are increasingly using LLMs to query the scientific literature and spot patterns in experimental data. 

But it’s still early days in turning those AI tools into actual materials discoveries. The use of AI to run autonomous labs, in particular, is just getting underway; making and testing stuff takes time and lots of money. The morning I visited Lila Sciences, its labs were largely empty, and it’s now preparing to move into a much larger space a few miles away. Periodic Labs is just beginning to set up its lab in San Francisco. It’s starting with manual synthesis guided by AI predictions; its robotic high-throughput lab will come soon. Radical AI reports that its lab is almost fully autonomous but plans to soon move to a larger space.

When I talk to the scientific founders of these startups, I hear a renewed excitement about a field that long operated in the shadows of drug discovery and genomic medicine. For one thing, there is the money. “You see this enormous enthusiasm to put AI and materials together,” says Ceder. “I’ve never seen this much money flow into materials.”

Reviving the materials industry is a challenge that goes beyond scientific advances, however. It means selling companies on a whole new way of doing R&D.

But the startups benefit from a huge dose of confidence borrowed from the rest of the AI industry. And maybe that, after years of playing it safe, is just what the materials business needs.

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology.

Expanded carrier screening: Is it worth it?

Carrier screening  tests would-be parents for hidden genetic mutations that might affect their children. It initially involved testing for specific genes in at-risk populations.

Expanded carrier screening takes things further, giving would-be parents an option to test for a wide array of diseases in prospective parents and egg and sperm donors.

The companies offering these screens “started out with 100 genes, and now some of them go up to 2,000,” Sara Levene, genetics counsellor at Guided Genetics, said at a meeting I attended this week. “It’s becoming a bit of an arms race amongst labs, to be honest.”

But expanded carrier screening comes with downsides. And it isn’t for everyone. Read the full story.

—Jessica Hamzelou

This article first appeared in The Checkup, MIT Technology Review’s weekly biotech newsletter. To receive it in your inbox every Thursday, and read articles like this first, sign up here.

Southeast Asia seeks its place in space

It’s a scorching October day in Bangkok and I’m wandering through the exhibits at the Thai Space Expo, held in one of the city’s busiest shopping malls, when I do a double take. Amid the flashy space suits and model rockets on display, there’s a plain-looking package of Thai basil chicken. I’m told the same kind of vacuum-­sealed package has just been launched to the International Space Station.

It’s an unexpected sight, one that reflects the growing excitement within the Southeast Asian space sector. And while there is some uncertainty about how exactly the region’s space sector may evolve, there is plenty of optimism, too. Read the full story.

—Jonathan O’Callaghan

This story is from the next print issue of MIT Technology Review magazine. If you haven’t already, subscribe now to receive future issues once they land.

The must-reads

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

1 Disney just signed a major deal with OpenAI
Meaning you’ll soon be able to create Sora clips starring 200 Marvel, Pixel and Star Wars characters. (Hollywood Reporter $)
+ Disney used to be openly skeptical of AI. What changed? (WSJ $)
+ It’s not feeling quite so friendly towards Google, however. (Ars Technica)
+ Expect a load of AI slop making its way to Disney Plus. (The Verge)

2 Donald Trump has blocked US states from enforcing their own AI rules
But technically, only Congress has the power to override state laws. (NYT $)
+ A new task force will seek out states with “inconsistent” AI rules. (Engadget)
+ The move is particularly bad news for California. (The Markup)

3 Reddit is challenging Australia’s social media ban for teens
It’s arguing that the ban infringes on their freedom of political communication. (Bloomberg $)
+ We’re learning more about the mysterious machinations of the teenage brain. (Vox)

4 ChatGPT’s “adult mode” is due to launch early next year
But OpenAI admits it needs to improve its age estimation tech first. (The Verge)
+ It’s pretty easy to get DeepSeek to talk dirty. (MIT Technology Review)

5 The death of Running Tide’s carbon removal dream
The company’s demise is a wake-up call to others dabbling in experimental tech. (Wired $)
+ We first wrote about Running Tide’s issues back in 2022. (MIT Technology Review)
+ What’s next for carbon removal? (MIT Technology Review)

6 That dirty-talking AI teddy bear wasn’t a one-off
It turns out that a wide range of LLM-powered toys aren’t suitable for children. (NBC News)
+ AI toys are all the rage in China—and now they’re appearing on shelves in the US too. (MIT Technology Review)

7 These are the cheapest places to create a fake online account
For a few cents, scammers can easily set up bots. (FT $)

8 How professors are attempting to AI-proof exams
ChatGPT won’t help you cut corners to ace an oral examination. (WP $)

9 Can a font be woke?
Marco Rubio seems to think so. (The Atlantic $)

10 Next year is all about maximalist circus decor
That’s according to Pinterest’s trend predictions for 2026. (The Guardian)

Quote of the day

 “Trump is delivering exactly what his billionaire benefactors demanded—all at the expense of our kids, our communities, our workers, and our planet.” 

—Senator Ed Markey criticizes Donald Trump’s decision to sign an order cracking down on US states’ ability to self-regulate AI, the Wall Street Journal reports.

One more thing

Taiwan’s “silicon shield” could be weakening

Taiwanese politics increasingly revolves around one crucial question: Will China invade? China’s ruling party has wanted to seize Taiwan for more than half a century. But in recent years, China’s leader, Xi Jinping, has placed greater emphasis on the idea of “taking back” the island (which the Chinese Communist Party, or CCP, has never controlled).

Many in Taiwan and elsewhere think one major deterrent has to do with the island’s critical role in semiconductor manufacturing. Taiwan produces the majority of the world’s semiconductors and more than 90% of the most advanced chips needed for AI applications.

But now some Taiwan specialists and some of the island’s citi­zens are worried that this “silicon shield,” if it ever existed, is cracking. Read the full story.

—Johanna M. Costigan

We can still have nice things

A place for comfort, fun and distraction to brighten up your day. (Got any ideas? Drop me a line or skeet ’em at me.)

+ Reasons to be cheerful: people are actually nicer than we think they are.
+ This year’s Krampus Run in Whitby—the Yorkshire town that inspired Bram Stoker’s Dracula—looks delightfully spooky.
+ How to find the magic in that most mundane of locations: the airport.
+ The happiest of birthdays to Dionne Warwick, who turns 85 today.

It’s a scorching October day in Bangkok and I’m wandering through the exhibits at the Thai Space Expo, held in one of the city’s busiest shopping malls, when I do a double take. Amid the flashy space suits and model rockets on display, there’s a plain-looking package of Thai basil chicken. I’m told the same kind of vacuum-­sealed package has just been launched to the International Space Station.

“This is real chicken that we sent to space,” says a spokesperson for the business behind the stunt, Charoen Pokphand Foods, the biggest food company in Thailand.

It’s an unexpected sight, one that reflects the growing excitement within the Southeast Asian space sector. At the expo, held among designer shops and street-food stalls, enthusiastic attendees have converged from emerging space nations such as Vietnam, Malaysia, Singapore, and of course Thailand to showcase Southeast Asia’s fledgling space industry.

While there is some uncertainty about how exactly the region’s space sector may evolve, there is plenty of optimism, too. “Southeast Asia is perfectly positioned to take leadership as a space hub,” says Candace Johnson, a partner in Seraphim Space, a UK investment firm that operates in Singapore. “There are a lot of opportunities.”

For example, Thailand may build a spaceport to launch rockets in the next few years, the country’s Geo-Informatics and Space Technology Development Agency announced the day before the expo started. “We don’t have a spaceport in Southeast Asia,” says Atipat Wattanuntachai, acting head of the space economy advancement division at the agency. “We saw a gap.” Because Thailand is so close to the equator, those rockets would get an additional boost from Earth’s rotation.

All kinds of companies here are exploring how they might tap into the global space economy. VegaCosmos, a startup based in Hanoi, Vietnam, is looking at ways to use satellite data for urban planning. The Electricity Generating Authority of Thailand is monitoring rainstorms from space to predict landslides. And the startup Spacemap, from Seoul, South Korea, is developing a new tool to better track satellites in orbit, which the US Space Force has invested in.

It’s the space chicken that caught my eye, though, perhaps because it reflects the juxtaposition of tradition and modernity seen across Bangkok, a city of ancient temples nestled next to glittering skyscrapers.

In June, astronauts on the space station were treated to this popular dish, known as pad krapow. It’s more commonly served up by street vendors, but this time it was delivered on a private mission operated by the US-based company Axiom Space. Charoen Pokphand is now using the stunt to say its chicken is good enough for NASA (sadly, I wasn’t able to taste it to weigh in).

Other Southeast Asian industries could also lend expertise to future space missions. Johnson says the region could leverage its manufacturing prowess to develop better semiconductors for satellites, for example, or break into the in-space manufacturing market.

I left the expo on a Thai longboat down the Chao Phraya River that weaves through Bangkok, with visions of astronauts tucking into some pad krapow in my head and imagining what might come next.

Jonathan O’Callaghan is a freelance space journalist based in Bangkok who covers commercial spaceflight, astrophysics, and space exploration.

This week I’ve been thinking about babies. Healthy ones. Perfect ones. As you may have read last week, my colleague Antonio Regalado came face to face with a marketing campaign in the New York subway asking people to “have your best baby.”

The company behind that campaign, Nucleus Genomics, says it offers customers a way to select embryos for a range of traits, including height and IQ. It’s an extreme proposition, but it does seem to be growing in popularity—potentially even in the UK, where it’s illegal.

The other end of the screening spectrum is transforming too. Carrier screening, which tests would-be parents for hidden genetic mutations that might affect their children, initially involved testing for specific genes in at-risk populations.

Now, it’s open to almost everyone who can afford it. Companies will offer to test for hundreds of genes to help people make informed decisions when they try to become parents. But expanded carrier screening comes with downsides. And it isn’t for everyone.

That’s what I found earlier this week when I attended the Progress Educational Trust’s annual conference in London.

First, a bit of background. Our cells carry 23 pairs of chromosomes, each with thousands of genes. The same gene—say, one that codes for eye color—can come in different forms, or alleles. If the allele is dominant, you only need one copy to express that trait. That’s the case for the allele responsible for brown eyes. 

If the allele is recessive, the trait doesn’t show up unless you have two copies. This is the case with the allele responsible for blue eyes, for example.

Things get more serious when we consider genes that can affect a person’s risk of disease. Having a single recessive disease-causing gene typically won’t cause you any problems. But a genetic disease could show up in children who inherit the same recessive gene from both parents. There’s a 25% chance that two “carriers” will have an affected child. And those cases can come as a shock to the parents, who tend to have no symptoms and no family history of disease.

This can be especially problematic in communities with high rates of those alleles. Consider Tay-Sachs disease—a rare and fatal neurodegenerative disorder caused by a recessive genetic mutation. Around one in 25 members of the Ashkenazi Jewish population is a healthy carrier for Tay-Sachs. Screening would-be parents for those recessive genes can be helpful. Carrier screening efforts in the Jewish community, which have been running since the 1970s, have massively reduced cases of Tay-Sachs.

Expanded carrier screening takes things further. Instead of screening for certain high-risk alleles in at-risk populations, there’s an option to test for a wide array of diseases in prospective parents and egg and sperm donors. The companies offering these screens “started out with 100 genes, and now some of them go up to 2,000,” Sara Levene, genetics counsellor at Guided Genetics, said at the meeting. “It’s becoming a bit of an arms race amongst labs, to be honest.”

There are benefits to expanded carrier screening. In most cases, the results are reassuring. And if something is flagged, prospective parents have options; they can often opt for additional testing to get more information about a particular pregnancy, for example, or choose to use other donor eggs or sperm to get pregnant. But there are also downsides. For a start, the tests can’t entirely rule out the risk of genetic disease.

Earlier this week, the BBC reported news of a sperm donor who had unwittingly passed on to at least 197 children in Europe a genetic mutation that dramatically increased the risk of cancer. Some of those children have already died.

It’s a tragic case. That donor had passed screening checks. The (dominant) mutation appears to have occurred in his testes, affecting around 20% of his sperm. It wouldn’t have shown up in a screen for recessive alleles, or even a blood test.

Even recessive diseases can be influenced by many genes, some of which won’t be included in the screen. And the screens don’t account for other factors that could influence a person’s risk of disease, such as epigenetics, microbiome, or even lifestyle.

“There’s always a 3% to 4% chance [of having] a child with a medical issue regardless of the screening performed,” said Jackson Kirkman-Brown, professor of reproductive biology at the University of Birmingham, at the meeting.

The tests can also cause stress. As soon as a clinician even mentions expanded carrier screening, it adds to the mental load of the patient, said Kirkman-Brown: “We’re saying this is another piece of information you need to worry about.”

People can also feel pressured to undergo expanded carrier screening even when they are ambivalent about it, said Heidi Mertes, a medical ethicist at Ghent University. “Once the technology is there, people feel like if they don’t take this opportunity up, then they are kind of doing something wrong or missing out,” she said.

My takeaway from the presentations was that while expanded carrier screening can be useful, especially for people from populations with known genetic risks, it won’t be for everyone.

I also worry that, as with the genetic tests offered by Nucleus, its availability gives the impression that it is possible to have a “perfect” baby—even if that only means “free from disease.” The truth is that there’s a lot about reproduction that we can’t control.

The decision to undergo expanded carrier screening is a personal choice. But as Mertes noted at the meeting: “Just because you can doesn’t mean you should.”

This article first appeared in The Checkup, MIT Technology Review’s weekly biotech newsletter. To receive it in your inbox every Thursday, and read articles like this first, sign up here.

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology.

Solar geoengineering startups are getting serious

Solar geoengineering aims to manipulate the climate by bouncing sunlight back into space. In theory, it could ease global warming. But as interest in the idea grows, so do concerns about potential consequences.

A startup called Stardust Solutions recently raised a $60 million funding round, the largest known to date for a geoengineering startup. My colleague James Temple has a new story out about the company, and how its emergence is making some researchers nervous.

So far, the field has been limited to debates, proposed academic research, and—sure—a few fringe actors to keep an eye on. Now things are getting more serious. So what does it mean for geoengineering, and for the climate? Read the full story.

—Casey Crownhart

This article is from The Spark, MIT Technology Review’s weekly climate newsletter. To receive it in your inbox every Wednesday, sign up here.

If you’re interested in reading more about solar geoengineering, check out:

+ Why the for-profit race into solar geoengineering is bad for science and public trust. Read the full story.

+ Why we need more research—including outdoor experiments—to make better-informed decisions about such climate interventions.

+ The hard lessons of Harvard’s failed geoengineering experiment, which was officially terminated last year. Read the full story.

+ How this London nonprofit became one of the biggest backers of geoengineering research.

+ The technology could alter the entire planet. These groups want every nation to have a say.

The must-reads

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

1 OpenAI is being sued for wrongful death
By the estate of a woman killed by her son after he engaged in delusion-filled conversations with ChatGPT. (WSJ $)
+ The chatbot appeared to validate Stein-Erik Soelberg’s conspiratorial ideas. (WP $)
+ It’s the latest in a string of wrongful death legal actions filed against chatbot makers. (ABC News)

2 ICE is tracking pregnant immigrants through specifically-developed smartwatches
They’re unable to take the devices off, even during labor. (The Guardian)
+ Pregnant and postpartum women say they’ve been detained in solitary confinement. (Slate $)
+ Another effort to track ICE raids has been taken offline. (MIT Technology Review)

3 Meta’s new AI hires aren’t making friends with the rest of the company
Tensions are rife between the AGI team and other divisions. (NYT $)
+ Mark Zuckerberg is keen to make money off the company’s AI ambitions. (Bloomberg $)
+ Meanwhile, what’s life like for the remaining Scale AI team? (Insider $)

4 Google DeepMind is building its first materials science lab in the UK
It’ll focus on developing new materials to build superconductors and solar cells. (FT $) 

5 The new space race is to build orbital data centers
And Blue Origin is winning, apparently. (WSJ $)
+ Plenty of companies are jostling for their slice of the pie. (The Verge)
+ Should we be moving data centers to space? (MIT Technology Review)

6 Inside the quest to find out what causes Parkinson’s
A growing body of work suggests it may not be purely genetic after all. (Wired $)

7 Are you in TikTok’s cat niche? 
If so, you’re likely to be in these other niches too. (WP $)

8 Why do our brains get tired?
Researchers are trying to get to the bottom of it.  (Nature $)

9 Microsoft’s boss has built his own cricket app
Satya Nadella can’t get enough of the sound of leather on willow. (Bloomberg $)

10 How much vibe coding is too much vibe coding? 
One journalist’s journey into the heart of darkness. (Rest of World)
+ What is vibe coding, exactly? (MIT Technology Review)

Quote of the day

“I feel so much pain seeing his sad face…I hope for a New Year’s miracle.”

—A child in Russia sends a message to the Kremlin-aligned Safe Internet League explaining the impact of the country’s decision to block access to the wildly popular gaming platform Roblox on their brother, the Washington Post reports.

 One more thing

Why it’s so hard to stop tech-facilitated abuse

After Gioia had her first child with her then husband, he installed baby monitors throughout their home—to “watch what we were doing,” she says, while he went to work. She’d turn them off; he’d get angry. By the time their third child turned seven, Gioia and her husband had divorced, but he still found ways to monitor her behavior. 

One Christmas, he gave their youngest a smartwatch. Gioia showed it to a tech-savvy friend, who found that the watch had a tracking feature turned on. It could be turned off only by the watch’s owner—her ex.

Gioia is far from alone. In fact, tech-facilitated abuse now occurs in most cases of intimate partner violence—and we’re doing shockingly little to prevent it. Read the full story. 

—Jessica Klein

We can still have nice things

A place for comfort, fun and distraction to brighten up your day. (Got any ideas? Drop me a line or skeet ’em at me.)

+ The New Yorker has picked its best TV shows of 2025. Let the debate commence!
+ Check out the winners of this year’s Drone Photo Awards.
+ I’m sorry to report you aren’t half as intuitive as you think you are when it comes to deciphering your dog’s emotions.
+ Germany’s “home of Christmas” sure looks magical.

Solar geoengineering aims to manipulate the climate by bouncing sunlight back into space. In theory, it could ease global warming. But as interest in the idea grows, so do concerns about potential consequences.

A startup called Stardust Solutions recently raised a $60 million funding round, the largest known to date for a geoengineering startup. My colleague James Temple has a new story out about the company, and how its emergence is making some researchers nervous.

So far, the field has been limited to debates, proposed academic research, and—sure—a few fringe actors to keep an eye on. Now things are getting more serious. What does it mean for geoengineering, and for the climate?

Researchers have considered the possibility of addressing planetary warming this way for decades. We already know that volcanic eruptions, which spew sulfur dioxide into the atmosphere, can reduce temperatures. The thought is that we could mimic that natural process by spraying particles up there ourselves.

The prospect is a controversial one, to put it lightly. Many have concerns about unintended consequences and uneven benefits. Even public research led by top institutions has faced barriers—one famous Harvard research program was officially canceled last year after years of debate.

One of the difficulties of geoengineering is that in theory a single entity, like a startup company, could make decisions that have a widespread effect on the planet. And in the last few years, we’ve seen more interest in geoengineering from the private sector. 

Three years ago, James broke the story that Make Sunsets, a California-based company, was already releasing particles into the atmosphere in an effort to tweak the climate.

The company’s CEO Luke Iseman went to Baja California in Mexico, stuck some sulfur dioxide into a weather balloon, and sent it skyward. The amount of material was tiny, and it’s not clear that it even made it into the right part of the atmosphere to reflect any sunlight.

But fears that this group or others could go rogue and do their own geoengineering led to widespread backlash. Mexico announced plans to restrict geoengineering experiments in the country a few weeks after that news broke.

You can still buy cooling credits from Make Sunsets, and the company was just granted a patent for its system. But the startup is seen as something of a fringe actor.

Enter Stardust Solutions. The company has been working under the radar for a few years, but it has started talking about its work more publicly this year. In October, it announced a significant funding round, led by some top names in climate investing. “Stardust is serious, and now it’s raised serious money from serious people,” as James puts it in his new story.

That’s making some experts nervous. Even those who believe we should be researching geoengineering are concerned about what it means for private companies to do so.

“Adding business interests, profit motives, and rich investors into this situation just creates more cause for concern, complicating the ability of responsible scientists and engineers to carry out the work needed to advance our understanding,” write David Keith and Daniele Visioni, two leading figures in geoengineering research, in a recent opinion piece for MIT Technology Review.

Stardust insists that it won’t move forward with any geoengineering until and unless it’s commissioned to do so by governments and there are rules and bodies in place to govern use of the technology.

But there’s no telling how financial pressure might change that, down the road. And we’re already seeing some of the challenges faced by a private company in this space: the need to keep trade secrets.

Stardust is currently not sharing information about the particles it intends to release into the sky, though it says it plans to do so once it secures a patent, which could happen as soon as next year. The company argues that its proprietary particles will be safe, cheap to manufacture, and easier to track than the already abundant sulfur dioxide. But at this point, there’s no way for external experts to evaluate those claims.

As Keith and Visioni put it: “Research won’t be useful unless it’s trusted, and trust depends on transparency.”

This article is from The Spark, MIT Technology Review’s weekly climate newsletter. To receive it in your inbox every Wednesday, sign up here.

In this exclusive subscriber-only eBook, you’ll learn about a new method that scientists have uncovered to look at the ways our bodies are aging.

by  Jessica Hamzelou October 14, 2025

Table of Contents:

• Clocks kick off
• Black-box clocks
• How to be young again
• Dogs and dolphins
• When young meets old

Related Stories:

• How aging clocks can help us understand why we age—and if we can reverse it
• The quest to legitimize longevity medicine

Access all subscriber-only eBooks:

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology.

How one controversial startup hopes to cool the planet

Stardust Solutions believes that it can solve climate change—for a price. 

The Israel-based geoengineering startup has said it expects nations will soon pay it more than a billion dollars a year to launch specially equipped aircraft into the stratosphere. Once they’ve reached the necessary altitude, those planes will disperse particles engineered to reflect away enough sunlight to cool down the planet, purportedly without causing environmental side effects. 

But numerous solar geoengineering researchers are skeptical that Stardust will line up the customers it needs to carry out a global deployment in the next decade. They’re also highly critical of the idea of a private company setting the global temperature for us. Read the full story.

—James Temple

MIT Technology Review Narrated: Is this the electric grid of the future?  

In Nebraska, a publicly owned utility company is tackling the challenges of delivering on reliability, affordability, and sustainability. It aims to reach net zero by 2040—here’s how it plans to get there.

This is our latest story to be turned into a MIT Technology Review Narrated podcast, which we’re publishing each week on Spotify and Apple Podcasts. Just navigate to MIT Technology Review Narrated on either platform, and follow us to get all our new content as it’s released.

The must-reads

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

1 Australia’s social media ban for teens has just come into force
The whole world will be watching to see what happens next. (The Guardian)
+ Opinions about the law are sharply divided among Australians. (BBC)
+ Plenty of teens hate it, naturally. (WP $)
+ A third of US teens are on their phones “almost constantly.” (NYT $)

2 This has been the second-hottest year since records began
Mean temperatures approached 1.5°C above the preindustrial average. (New Scientist $)
+ Meanwhile world leaders at this year’s UN climate talks couldn’t even agree to use the phrase ‘fossil fuels’ in the final draft. (MIT Technology Review)

3 OpenAI is in trouble
It’s rapidly losing its technological edge to competitors like Google and Anthropic. (The Atlantic $)
+ Silicon Valley is working harder than ever to sell AI to us. (Wired $)
+ There’s a new industry-wide push to agree shared standards for AI agents. (TechCrunch)
+ No one can explain how AI really works—not even the experts attending AI’s biggest research gathering. (NBC)

4 MAGA influencers want Trump to kill the Netflix/Warner Bros deal
They argue Netflix is simply too woke (after all, it employs the Obamas.) (WP $)

5 AI slop videos have taken over social media
It’s now almost impossible to tell if what you’re seeing is real or not. (NYT $)

6 Trump’s system to weed out noncitizen voters is flagging US citizens 
Once alerted, people have 30 days to provide proof of citizenship before they lose their ability to vote. (NPR)
+ The US is planning to ask visitors to disclose five years of social media history. (WP $)
+ How open source voting machines could boost trust in US elections. (MIT Technology Review)

7 Virtual power plants are having a moment
Here’s why they’re poised to play a significant role in meeting energy demand over the next decade. (IEEE Spectrum)
+ How virtual power plants are shaping tomorrow’s energy system. (MIT Technology Review)

8 New devices are about to get (even) more expensive
You can thank AI for pushing up the price of RAM for the rest of us. (The Verge)

9 People hated the McDonald’s AI ad so much the company pulled it 
How are giant corporations still falling into this exact trap every holiday season? (Forbes)  

10 Why is ice slippery? There’s a new hypothesis 
You might think you know. But it’s still fiercely debated among ice researchers! (Quanta $)

Quote of the day

“We’re pleased to be the first, we’re proud to be the first, and we stand ready to help any other jurisdiction who seeks to do these things.”

—Australia’s communications minister Anika Wells tells the BBC how she feels about her government’s decision to ban social media for under-16s. 

One more thing

The entrepreneur dreaming of a factory of unlimited organs

At any given time, the US transplant waiting list is about 100,000 people long. Thousands die waiting, and many more never make the list to begin with. Entrepreneur Martine Rothblatt wants to address this by growing organs compatible with human bodies in genetically modified pigs.

In recent years, US doctors have attempted seven pig-to-human transplants, the most dramatic of which was a case where a 57-year-old man with heart failure lived two months with a pig heart supplied by Rothblatt’s company. 

The experiment demonstrated the first life-sustaining pig-to-human organ transplant—and paved the way towards an organized clinical trial to prove they save lives consistently. Read the full story.

—Antonio Regalado

We can still have nice things

A place for comfort, fun and distraction to brighten up your day. (Got any ideas? Drop me a line or skeet ’em at me.)

+ I want to eat all of these things, starting with the hot chocolate cookies. 
+ Even one minute is enough time to enjoy some of the benefits of mindfulness.
+ The Geminid meteor shower will reach its peak this weekend. Here’s how to see it. 
+ I really enjoy Leah Gardner’s still life paintings.

At a regional hospital, a cardiac patient’s lab results sit behind layers of encryption, accessible to his surgeon but shielded from those without strictly need-to-know status. Across the street at a credit union, a small business owner anxiously awaits the all-clear for a wire transfer, unaware that fraud detection systems have flagged it for further review.

Such scenarios illustrate how companies in regulated industries juggle competing directives: Move data and process transactions quickly enough to save lives and support livelihoods, but carefully enough to maintain ironclad security and satisfy regulatory scrutiny.

Organizations subject to such oversight walk a fine line every day. And recently, a number of curveballs have thrown off that hard-won equilibrium. Agencies are ramping up oversight thanks to escalating data privacy concerns; insurers are tightening underwriting and requiring controls like MFA and privileged-access governance as a condition of coverage. Meanwhile, the shifting VMware landscape has introduced more complexity for IT teams tasked with planning long-term infrastructure strategies. 

Download the full article

This content was produced by Insights, the custom content arm of MIT Technology Review. It was not written by MIT Technology Review’s editorial staff.

This content was researched, designed, and written by human writers, editors, analysts, and illustrators. This includes the writing of surveys and collection of data for surveys. AI tools that may have been used were limited to secondary production processes that passed thorough human review.

Stardust Solutions believes that it can solve climate change—for a price.

The Israel-based geoengineering startup has said it expects  nations will soon pay it more than a billion dollars a year to launch specially equipped aircraft into the stratosphere. Once they’ve reached the necessary altitude, those planes will disperse particles engineered to reflect away enough sunlight to cool down the planet, purportedly without causing environmental side effects. 

The proprietary (and still secret) particles could counteract all the greenhouse gases the world has emitted over the last 150 years, the company stated in a 2023 pitch deck it presented to venture capital firms. In fact, it’s the “only technologically feasible solution” to climate change, the company said.

The company disclosed it raised $60 million in funding in October, marking by far the largest known funding round to date for a startup working on solar geoengineering.

Stardust is, in a sense, the embodiment of Silicon Valley’s simmering frustration with the pace of academic research on the technology. It’s a multimillion-dollar bet that a startup mindset can advance research and development that has crept along amid scientific caution and public queasiness.

But numerous researchers focused on solar geoengineering are deeply skeptical that Stardust will line up the government customers it would need to carry out a global deployment as early as 2035, the plan described in its earlier investor materials—and aghast at the suggestion that it ever expected to move that fast. They’re also highly critical of the idea that a company would take on the high-stakes task of setting the global temperature, rather than leaving it to publicly funded research programs.

“They’ve ignored every recommendation from everyone and think they can turn a profit in this field,” says Douglas MacMartin, an associate professor at Cornell University who studies solar geoengineering. “I think it’s going to backfire. Their investors are going to be dumping their money down the drain, and it will set back the field.”

The company has finally emerged from stealth mode after completing its funding round, and its CEO, Yanai Yedvab, agreed to conduct one of the company’s first extensive interviews with MIT Technology Review for this story.

Yedvab walked back those ambitious projections a little, stressing that the actual timing of any stratospheric experiments, demonstrations, or deployments will be determined by when governments decide it’s appropriate to carry them out. Stardust has stated clearly that it will move ahead with solar geoengineering only if nations pay it to proceed, and only once there are established rules and bodies guiding the use of the technology.

That decision, he says, will likely be dictated by how bad climate change becomes in the coming years.

“It could be a situation where we are at the place we are now, which is definitely not great,” he says. “But it could be much worse. We’re saying we’d better be ready.”

“It’s not for us to decide, and I’ll say humbly, it’s not for these researchers to decide,” he adds. “It’s the sense of urgency that will dictate how this will evolve.”

The building blocks

No one is questioning the scientific credentials of Stardust. The company was founded in 2023 by a trio of prominent researchers, including Yedvab, who served as deputy chief scientist at the Israeli Atomic Energy Commission. The company’s lead scientist, Eli Waxman, is the head of the department of particle physics and astrophysics at the Weizmann Institute of Science. Amyad Spector, the chief product officer, was previously a nuclear physicist at Israel’s secretive Negev Nuclear Research Center.

Stardust says it employs 25 scientists, engineers, and academics. The company is based in Ness Ziona, Israel, and plans to open a US headquarters soon. 

Yedvab says the motivation for starting Stardust was simply to help develop an effective means of addressing climate change. 

“Maybe something in our experience, in the tool set that we bring, can help us in contributing to solving one of the greatest problems humanity faces,” he says.

Lowercarbon Capital, the climate-tech-focused investment firm  cofounded by the prominent tech investor Chris Sacca, led the $60 million investment round. Future Positive, Future Ventures, and Never Lift Ventures, among others, participated as well.

AWZ Ventures, a firm focused on security and intelligence technologies, co-led the company’s earlier seed round, which totaled $15 million.

Yedvab says the company will use that money to advance research, development, and testing for the three components of its system, which are also described in the pitch deck: safe particles that could be affordably manufactured; aircraft dispersion systems; and a means of tracking particles and monitoring their effects.

“Essentially, the idea is to develop all these building blocks and to upgrade them to a level that will allow us to give governments the tool set and all the required information to make decisions about whether and how to deploy this solution,” he says. 

The company is, in many ways, the opposite of Make Sunsets, the first company that came along offering to send particles into the stratosphere—for a fee—by pumping sulfur dioxide into weather balloons and hand-releasing them into the sky. Many researchers viewed it as a provocative, unscientific, and irresponsible exercise in attention-gathering. 

But Stardust is serious, and now it’s raised serious money from serious people—all of which raises the stakes for the solar geoengineering field and, some fear, increases the odds that the world will eventually put the technology to use.

“That marks a turning point in that these types of actors are not only possible, but are real,” says Shuchi Talati, executive director of the Alliance for Just Deliberation on Solar Geoengineering, a nonprofit that strives to ensure that developing nations are included in the global debate over such climate interventions. “We’re in a more dangerous era now.”

Many scientists studying solar geoengineering argue strongly that universities, governments, and transparent nonprofits should lead the work in the field, given the potential dangers and deep public concerns surrounding a tool with the power to alter the climate of the planet. 

It’s essential to carry out the research with appropriate oversight, explore the potential downsides of these approaches, and publicly publish the results “to ensure there’s no bias in the findings and no ulterior motives in pushing one way or another on deployment or not,” MacMartin says. “[It] shouldn’t be foisted upon people without proper and adequate information.”

He criticized, for instance, the company’s claims to have developed what he described as their “magic aerosol particle,” arguing that the assertion that it is perfectly safe and inert can’t be trusted without published findings. Other scientists have also disputed those scientific claims.

Plenty of other academics say solar geoengineering shouldn’t be studied at all, fearing that merely investigating it starts the world down a slippery slope toward its use and diminishes the pressures to cut greenhouse-gas emissions. In 2022, hundreds of them signed an open letter calling for a global ban on the development and use of the technology, adding the concern that there is no conceivable way for the world’s nations to pull together to establish rules or make collective decisions ensuring that it would be used in “a fair, inclusive, and effective manner.”

“Solar geoengineering is not necessary,” the authors wrote. “Neither is it desirable, ethical, or politically governable in the current context.”

The for-profit decision 

Stardust says it’s important to pursue the possibility of solar geoengineering because the dangers of climate change are accelerating faster than the world’s ability to respond to it, requiring a new “class of solution … that buys us time and protects us from overheating.”

Yedvab says he and his colleagues thought hard about the right structure for the organization, finally deciding that for-profits working in parallel with academic researchers have delivered “most of the groundbreaking technologies” in recent decades. He cited advances in genome sequencing, space exploration, and drug development, as well as the restoration of the ozone layer.

He added that a for-profit structure was also required to raise funds and attract the necessary talent.

“There is no way we could, unfortunately, raise even a small portion of this amount by philanthropic resources or grants these days,” he says.

He adds that while academics have conducted lots of basic science in solar geoengineering, they’ve done very little in terms of building the technological capacities. Their geoengineering research is also primarily focused on the potential use of sulfur dioxide, because it is known to help reduce global temperatures after volcanic eruptions blast massive amounts of it into the stratospheric. But it has well-documented downsides as well, including harm to the protective ozone layer.

“It seems natural that we need better options, and this is why we started Stardust: to develop this safe, practical, and responsible solution,” the company said in a follow-up email. “Eventually, policymakers will need to evaluate and compare these options, and we’re confident that our option will be superior over sulfuric acid primarily in terms of safety and practicability.”

Public trust can be won not by excluding private companies, but by setting up regulations and organizations to oversee this space, much as the US Food and Drug Administration does for pharmaceuticals, Yedvab says.

“There is no way this field could move forward if you don’t have this governance framework, if you don’t have external validation, if you don’t have clear regulation,” he says.

Meanwhile, the company says it intends to operate transparently, pledging to publish its findings whether they’re favorable or not.

That will include finally revealing details about the particles it has developed, Yedvab says. 

Early next year, the company and its collaborators will begin publishing data or evidence “substantiating all the claims and disclosing all the information,” he says, “so that everyone in the scientific community can actually check whether we checked all these boxes.”

In the follow-up email, the company acknowledged that solar geoengineering isn’t a “silver bullet” but said it is “the only tool that will enable us to cool the planet in the short term, as part of a larger arsenal of technologies.”

“The only way governments could be in a position to consider [solar geoengineering] is if the work has been done to research, de-risk, and engineer safe and responsible solutions—which is what we see as our role,” the company added later. “We are hopeful that research will continue not just from us, but also from academic institutions, nonprofits, and other responsible companies that may emerge in the future.”

Ambitious projections

Stardust’s earlier pitch deck stated that the company expected to conduct its first “stratospheric aerial experiments” last year, though those did not move ahead (more on that in a moment).

On another slide, the company said it expected to carry out a “large-scale demonstration” around 2030 and proceed to a “global full-scale deployment” by about 2035. It said it expected to bring in roughly $200 million and $1.5 billion in annual revenue by those periods, respectively.

Every researcher interviewed for this story was adamant that such a deployment should not happen so quickly.

Given the global but uneven and unpredictable impacts of solar geoengineering, any decision to use the technology should be reached through an inclusive, global agreement, not through the unilateral decisions of individual nations, Talati argues. 

“We won’t have any sort of international agreement by that point given where we are right now,” she says.

A global agreement, to be clear, is a big step beyond setting up rules and oversight bodies—and some believe that such an agreement on a technology so divisive could never be achieved.

There’s also still a vast amount of research that must be done to better understand the negative side effects of solar geoengineering generally and any ecological impacts of Stardust’s materials specifically, adds Holly Buck, an associate professor at the University of Buffalo and author of After Geoengineering.

“It is irresponsible to talk about deploying stratospheric aerosol injection without fundamental research about its impacts,” Buck wrote in an email.

She says the timelines are also “unrealistic” because there are profound public concerns about the technology. Her polling work found that a significant fraction of the US public opposes even research (though polling varies widely). 

Meanwhile, most academic efforts to move ahead with even small-scale outdoor experiments have sparked fierce backlash. That includes the years-long effort by researchers then at Harvard to carry out a basic equipment test for their so-called ScopeX experiment. The high-altitude balloon would have launched from a flight center in Sweden, but the test was ultimately scratched amid objections from environmentalists and Indigenous groups. 

Given this baseline of public distrust, Stardust’s for-profit proposals only threaten to further inflame public fears, Buck says.

“I find the whole proposal incredibly socially naive,” she says. “We actually could use serious research in this field, but proposals like this diminish the chances of that happening.”

Those public fears, which cross the political divide, also mean politicians will see little to no political upside to paying Stardust to move ahead, MacMartin says.

“If you don’t have the constituency for research, it seems implausible to me that you’d turn around and give money to an Israeli company to deploy it,” he says.

An added risk is that if one nation or a small coalition forges ahead without broader agreement, it could provoke geopolitical conflicts. 

“What if Russia wants it a couple of degrees warmer, and India a couple of degrees cooler?” asked Alan Robock, a professor at Rutgers University, in the Bulletin of the Atomic Scientists in 2008. “Should global climate be reset to preindustrial temperature or kept constant at today’s reading? Would it be possible to tailor the climate of each region of the planet independently without affecting the others? If we proceed with geoengineering, will we provoke future climate wars?”

Revised plans

Yedvab says the pitch deck reflected Stardust’s strategy at a “very early stage in our work,” adding that their thinking has “evolved,” partly in response to consultations with experts in the field.

He says that the company will have the technological capacity to move ahead with demonstrations and deployments on the timelines it laid out but adds, “That’s a necessary but not sufficient condition.”

“Governments will need to decide where they want to take it, if at all,” he says. “It could be a case that they will say ‘We want to move forward.’ It could be a case that they will say ‘We want to wait a few years.’”

“It’s for them to make these decisions,” he says.

Yedvab acknowledges that the company has conducted flights in the lower atmosphere to test its monitoring system, using white smoke as a simulant for its particles, as the Wall Street Journal reported last year. It’s also done indoor tests of the dispersion system and its particles in a wind tunnel set up within its facility.

But in response to criticisms like the ones above, Yedvab says the company hasn’t conducted outdoor particle experiments and won’t move forward with them until it has approval from governments. 

“Eventually, there will be a need to conduct outdoor testing,” he says. “There is no way you can validate any solution without outdoor testing.” But such testing of sunlight reflection technology, he says, “should be done only working together with government and under these supervisions.”

Generating returns  

Stardust may be willing to wait for governments to be ready to deploy its system, but there’s no guarantee that its investors will have the same patience. In accepting tens of millions in venture capital, Stardust may now face financial pressures that could “drive the timelines,” says Gernot Wagner, a climate economist at Columbia University. 

And that raises a different set of concerns.

Obliged to deliver returns, the company might feel it must strive to convince government leaders that they should pay for its services, Talati says. 

“The whole point of having companies and investors is you want your thing to be used,” she says. “There’s a massive incentive to lobby countries to use it, and that’s the whole danger of having for-profit companies here.”

She argues those financial incentives threaten to accelerate the use of solar geoengineering ahead of broader international agreements and elevate business interests above the broader public good.

Stardust has “quietly begun lobbying on Capitol Hill” and has hired the law firm Holland & Knight, according to Politico.

It has also worked with Red Duke Strategies, a consulting firm based in McLean, Virginia, to develop “strategic relationships and communications that promote understanding and enable scientific testing,” according to a case study on the company’s  website. 

“The company needed to secure both buy-in and support from the United States government and other influential stakeholders to move forward,” Red Duke states. “This effort demanded a well-connected and authoritative partner who could introduce Stardust to a group of experts able to research, validate, deploy, and regulate its SRM technology.”

Red Duke didn’t respond to an inquiry from MIT Technology Review. Stardust says its work with the consulting firm was not a government lobbying effort.

Yedvab acknowledges that the company is meeting with government leaders in the US, Europe, its own region, and the Global South. But he stresses that it’s not asking any country to contribute funding or to sign off on deployments at this stage. Instead, it’s making the case for nations to begin crafting policies to regulate solar geoengineering.

“When we speak to policymakers—and we speak to policymakers; we don’t hide it—essentially, what we tell them is ‘Listen, there is a solution,’” he says. “‘It’s not decades away—it’s a few years away. And it’s your role as policymakers to set the rules of this field.’”

“Any solution needs checks and balances,” he says. “This is how we see the checks and balances.”

He says the best-case scenario is still a rollout of clean energy technologies that accelerates rapidly enough to drive down emissions and curb climate change.

“We are perfectly fine with building an option that will sit on the shelf,” he says. “We’ll go and do something else. We have a great team and are confident that we can find also other problems to work with.”

He says the company’s investors are aware of and comfortable with that possibility, supportive of the principles that will guide Stardust’s work, and willing to wait for regulations and government contracts.

Lowercarbon Capital didn’t respond to an inquiry from MIT Technology Review.

‘Sentiment of hope’

Others have certainly imagined the alternative scenario Yedvab raises: that nations will increasingly support the idea of geoengineering in the face of mounting climate catastrophes. 

In Kim Stanley Robinson’s 2020 novel, The Ministry for the Future, India unilaterally forges ahead with solar geoengineering following a heat wave that kills millions of people. 

Wagner sketched a variation on that scenario in his 2021 book, Geoengineering: The Gamble, speculating that a small coalition of nations might kick-start a rapid research and deployment program as an emergency response to escalating humanitarian crises. In his version, the Philippines offers to serve as the launch site after a series of super-cyclones batter the island nation, forcing millions from their homes. 

It’s impossible to know today how the world will react if one nation or a few go it alone, or whether nations could come to agreement on where the global temperature should be set. 

But the lure of solar geoengineering could become increasingly enticing as more and more nations endure mass suffering, starvation, displacement, and death.

“We understand that probably it will not be perfect,” Yedvab says. “We understand all the obstacles, but there is this sentiment of hope, or cautious hope, that we have a way out of this dark corridor we are currently in.”

“I think that this sentiment of hope is something that gives us a lot of energy to move on forward,” he adds.

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology.

The State of AI: A vision of the world in 2030  

There are huge gulfs of opinion when it comes to predicting the near-future impacts of generative AI. In one camp there are those who predict that over the next decade the impact of AI will exceed that of the Industrial Revolution—a 150-year period of economic and social upheaval so great that we still live in the world it wrought. 

At the other end of the scale we have team ‘Normal Technology’: experts who push back not only on these sorts of predictions but on their foundational worldview. That’s not how technology works, they argue.

Advances at the cutting edge may come thick and fast, but change across the wider economy, and society as a whole, moves at human speed. Widespread adoption of new technologies can be slow; acceptance slower. AI will be no different. What should we make of these extremes? 

Read the full conversation between MIT Technology Review’s senior AI editor Will Douglas Heaven and Tim Bradshaw, FT global tech correspondent, about where AI will go next, and what our world will look like in the next five years.

This is the final edition of The State of AI, a collaboration between the Financial Times and MIT Technology Review. Read the rest of the series, and if you want to keep up-to-date with what’s going on in the world of AI, sign up to receive our free Algorithm newsletter every Monday.

How AI is changing the economy

There’s a lot at stake when it comes to understanding how AI is changing the economy at large. What’s the right outlook to have? Join Mat Honan, editor in chief, David Rotman, editor at large, and Richard Waters, FT columnist, at 1pm ET today to hear them discuss what’s happening across industries and the market. Sign up now to be part of this exclusive subscriber-only event.

The must-reads

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

1 Trump says he’ll sign an order blocking states from regulating AI
But he’s facing a lot of pushback, including from members of his own party. (CNN)
+ The whole debacle can be traced back to congressional inaction. (Semafor)

2 Google’s new smart glasses are getting rave reviews 
You’ll be able to get your hands on a pair in 2026. Watch out, Apple and Meta. (Tech Radar)

3 Trump gave the go-ahead for Nvidia to sell powerful AI chips to China
The US gets a 25% cut of the sales—but what does it lose longer-term? (WP $)
+ And how much could China stand to gain? (NYT $)
+ How a top Chinese AI model overcame US sanctions. (MIT Technology Review)

4 America’s data center backlash is here
Republican and Democrat alike, local residents are sick of rapidly rising power bills. (Vox $)
+ More than 200 environmental groups are demanding a US-wide moratorium on new data centers. (The Guardian)
+ The data center boom in the desert. (MIT Technology Review)

5 A quarter of teens are turning to AI chatbots for mental health support
Given the lack of real-world help, can you really blame them? (The Guardian)
+ Therapists are secretly using ChatGPT. Clients are triggered. (MIT Technology Review)

6 ICEBlock is suing the US government over its App Store removal 
Its creator is arguing that the Department of Justice’s demands to Apple violated his First Amendment rights. (404 Media)
+ It’s one of a number of ICE-tracking initiatives to be pulled by tech platforms this year. (MIT Technology Review)

7 This band quit Spotify, but it’s been replaced by AI knockoffs
The platform seems to be struggling against the tide of slop. (Futurism) 
+ AI is coming for music, too. (MIT Technology Review)

8 Think you’re immune to online ads? Think again
If you’re scrolling on social media, you’re being sold to. Relentlessly. (The Verge $)

9 People really do not like Microsoft Copilot
It’s like Clippy all over again, except it’s even less avoidable. (Quartz $)

10 The longest solar eclipse for 100 years is coming
And we’ll only have to wait until 2027 to see it! (Wired $)

Quote of the day

“Governments and MPs are shooting themselves in the foot by pandering to tech giants, because that just tells young people that they don’t care about our future.”

—Adele Zeynep Walton, founding member of online safety campaign group Ctrl+Alt+Reclaim, tells The Guardian why young activists are taking matters into their own hands. 

One more thing

Inside the long quest to advance Chinese writing technology

Every second of every day, someone is typing in Chinese. Though the mechanics look a little different from typing in English—people usually type the pronunciation of a character and then pick it out of a selection that pops up, autocomplete-style—it’s hard to think of anything more quotidian. The software that allows this exists beneath the awareness of pretty much everyone who uses it. It’s just there.

What’s largely been forgotten is that a large cast of eccentrics and linguists, engineers and polymaths, spent much of the 20th century torturing themselves over how Chinese was ever going to move away from the ink brush to any other medium. Read the full story.

—Veronique Greenwood

We can still have nice things

A place for comfort, fun and distraction to brighten up your day. (Got any ideas? Drop me a line or skeet ’em at me.)

+ Pantone chose a ‘calming’ shade of white for its Color of 2026… and people are fuming. 
+ Ozempic needles on the Christmas tree, anyone? Here’s why we’re going crazy for weird baubles. 
+ Can relate to this baby seal for instinctively heading to the nearest pub.
+ Thrilled to see One Battle After Another get so many Golden Globes nominations.

Welcome back to The State of AI, a new collaboration between the Financial Times and MIT Technology Review. Every Monday, writers from both publications debate one aspect of the generative AI revolution reshaping global power. You can read the rest of the series here.

In this final edition, MIT Technology Review’s senior AI editor Will Douglas Heaven talks with Tim Bradshaw, FT global tech correspondent, about where AI will go next, and what our world will look like in the next five years.

(As part of this series, join MIT Technology Review’s editor in chief, Mat Honan, and editor at large, David Rotman, for an exclusive conversation with Financial Times columnist Richard Waters on how AI is reshaping the global economy. Live on Tuesday, December 9 at 1:00 p.m. ET. This is a subscriber-only event and you can sign up here.)

Will Douglas Heaven writes: 

Every time I’m asked what’s coming next, I get a Luke Haines song stuck in my head: “Please don’t ask me about the future / I am not a fortune teller.” But here goes. What will things be like in 2030? My answer: same but different. 

There are huge gulfs of opinion when it comes to predicting the near-future impacts of generative AI. In one camp we have the AI Futures Project, a small donation-funded research outfit led by former OpenAI researcher Daniel Kokotajlo. The nonprofit made a big splash back in April with AI 2027, a speculative account of what the world will look like two years from now. 

The story follows the runaway advances of an AI firm called OpenBrain (any similarities are coincidental, etc.) all the way to a choose-your-own-adventure-style boom or doom ending. Kokotajlo and his coauthors make no bones about their expectation that in the next decade the impact of AI will exceed that of the Industrial Revolution—a 150-year period of economic and social upheaval so great that we still live in the world it wrought.

At the other end of the scale we have team Normal Technology: Arvind Narayanan and Sayash Kapoor, a pair of Princeton University researchers and coauthors of the book AI Snake Oil, who push back not only on most of AI 2027’s predictions but, more important, on its foundational worldview. That’s not how technology works, they argue.

Advances at the cutting edge may come thick and fast, but change across the wider economy, and society as a whole, moves at human speed. Widespread adoption of new technologies can be slow; acceptance slower. AI will be no different. 

What should we make of these extremes? ChatGPT came out three years ago last month, but it’s still not clear just how good the latest versions of this tech are at replacing lawyers or software developers or (gulp) journalists. And new updates no longer bring the step changes in capability that they once did. 

And yet this radical technology is so new it would be foolish to write it off so soon. Just think: Nobody even knows exactly how this technology works—let alone what it’s really for. 

As the rate of advance in the core technology slows down, applications of that tech will become the main differentiator between AI firms. (Witness the new browser wars and the chatbot pick-and-mix already on the market.) At the same time, high-end models are becoming cheaper to run and more accessible. Expect this to be where most of the action is: New ways to use existing models will keep them fresh and distract people waiting in line for what comes next. 

Meanwhile, progress continues beyond LLMs. (Don’t forget—there was AI before ChatGPT, and there will be AI after it too.) Technologies such as reinforcement learning—the powerhouse behind AlphaGo, DeepMind’s board-game-playing AI that beat a Go grand master in 2016—is set to make a comeback. There’s also a lot of buzz around world models, a type of generative AI with a stronger grip on how the physical world fits together than LLMs display. 

Ultimately, I agree with team Normal Technology that rapid technological advances do not translate to economic or societal ones straight away. There’s just too much messy human stuff in the middle. 

But Tim, over to you. I’m curious to hear what your tea leaves are saying. 

Tim Bradshaw responds:

Will, I am more confident than you that the world will look quite different in 2030. In five years’ time, I expect the AI revolution to have proceeded apace. But who gets to benefit from those gains will create a world of AI haves and have-nots.

It seems inevitable that the AI bubble will burst sometime before the end of the decade. Whether a venture capital funding shakeout comes in six months or two years (I feel the current frenzy still has some way to run), swathes of AI app developers will disappear overnight. Some will see their work absorbed by the models upon which they depend. Others will learn the hard way that you can’t sell services that cost $1 for 50 cents without a firehose of VC funding.

How many of the foundation model companies survive is harder to call, but it already seems clear that OpenAI’s chain of interdependencies within Silicon Valley make it too big to fail. Still, a funding reckoning will force it to ratchet up pricing for its services.

When OpenAI was created in 2015, it pledged to “advance digital intelligence in the way that is most likely to benefit humanity as a whole.” That seems increasingly untenable. Sooner or later, the investors who bought in at a $500 billion price tag will push for returns. Those data centers won’t pay for themselves. By that point, many companies and individuals will have come to depend on ChatGPT or other AI services for their everyday workflows. Those able to pay will reap the productivity benefits, scooping up the excess computing power as others are priced out of the market.

Being able to layer several AI services on top of each other will provide a compounding effect. One example I heard on a recent trip to San Francisco: Ironing out the kinks in vibe coding is simply a matter of taking several passes at the same problem and then running a few more AI agents to look for bugs and security issues. That sounds incredibly GPU-intensive, implying that making AI really deliver on the current productivity promise will require customers to pay far more than most do today.

The same holds true in physical AI. I fully expect robotaxis to be commonplace in every major city by the end of the decade, and I even expect to see humanoid robots in many homes. But while Waymo’s Uber-like prices in San Francisco and the kinds of low-cost robots produced by China’s Unitree give the impression today that these will soon be affordable for all, the compute cost involved in making them useful and ubiquitous seems destined to turn them into luxuries for the well-off, at least in the near term.

The rest of us, meanwhile, will be left with an internet full of slop and unable to afford AI tools that actually work.

Perhaps some breakthrough in computational efficiency will avert this fate. But the current AI boom means Silicon Valley’s AI companies lack the incentives to make leaner models or experiment with radically different kinds of chips. That only raises the likelihood that the next wave of AI innovation will come from outside the US, be that China, India, or somewhere even farther afield.

Silicon Valley’s AI boom will surely end before 2030, but the race for global influence over the technology’s development—and the political arguments about how its benefits are distributed—seem set to continue well into the next decade. 

Will replies: 

I am with you that the cost of this technology is going to lead to a world of haves and have-nots. Even today, $200+ a month buys power users of ChatGPT or Gemini a very different experience from that of people on the free tier. That capability gap is certain to increase as model makers seek to recoup costs. 

We’re going to see massive global disparities too. In the Global North, adoption has been off the charts. A recent report from Microsoft’s AI Economy Institute notes that AI is the fastest-spreading technology in human history: “In less than three years, more than 1.2 billion people have used AI tools, a rate of adoption faster than the internet, the personal computer, or even the smartphone.” And yet AI is useless without ready access to electricity and the internet; swathes of the world still have neither. 

I still remain skeptical that we will see anything like the revolution that many insiders promise (and investors pray for) by 2030. When Microsoft talks about adoption here, it’s counting casual users rather than measuring long-term technological diffusion, which takes time. Meanwhile, casual users get bored and move on. 

How about this: If I live with a domestic robot in five years’ time, you can send your laundry to my house in a robotaxi any day of the week. 

JK! As if I could afford one. 

Further reading 

What is AI? It sounds like a stupid question, but it’s one that’s never been more urgent. In this deep dive, Will unpacks decades of spin and speculation to get to the heart of our collective technodream. 

AGI—the idea that machines will be as smart as humans—has hijacked an entire industry (and possibly the US economy). For MIT Technology Review’s recent New Conspiracy Age package, Will takes a provocative look at how AGI is like a conspiracy. 

The FT examined the economics of self-driving cars this summer, asking who will foot the multi-billion-dollar bill to buy enough robotaxis to serve a big city like London or New York.

A plausible counter-argument to Tim’s thesis on AI inequalities is that freely available open-source (or more accurately, “open weight”) models will keep pulling down prices. The US may want frontier models to be built on US chips but it is already losing the global south to Chinese software.

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology.

4 technologies that didn’t make our 2026 breakthroughs list

If you’re a longtime reader, you probably know that our newsroom selects 10 breakthroughs every year that we think will define the future. This group exercise is mostly fun and always engrossing, with plenty of lively discussion along the way, but at times it can also be quite difficult.  

The 2026 list will come out on January 12—so stay tuned. In the meantime, we wanted to share some of the technologies from this year’s reject pile, as a window into our decision-making process. These four technologies won’t be on our 2026 list of breakthroughs, but all were closely considered, and we think they’re worth knowing about. Read the full story to learn what they are. 

MIT Technology Review Narrated: The quest to find out how our bodies react to extreme temperatures 

Scientists hope to prevent deaths from climate change, but heat and cold are more complicated than we thought. Researchers around the world are revising rules about when extremes veer from uncomfortable to deadly. Their findings change how we should think about the limits of hot and cold—and how to survive in a new world. 

This is our latest story to be turned into a MIT Technology Review Narrated podcast, which we’re publishing each week on Spotify and Apple Podcasts. Just navigate to MIT Technology Review Narrated on either platform, and follow us to get all our new content as it’s released.

The must-reads

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

1 A CDC panel voted to recommend delaying the hepatitis B vaccine for babies
Overturning a 30-year policy that has contributed to a huge decline in the virus. (STAT)
+ Why childhood vaccines are a public health success story. (MIT Technology Review)

2 Critical climate risks are growing across the Arab region 
Drought is the most immediate problem countries are having to grapple with. (Ars Technica)
+ Why Tehran is running out of water. (Wired $)

3 Netflix is buying Warner Bros for $83 billion 
If approved, it’ll be one of the most significant mergers in Hollywood history. (NBC)
+ Trump says the deal “could be a problem” due to Netflix’s already huge market share. (BBC)

4 The EU is fining X $140 million 
For failing to comply with its new Digital Services Act. (NPR)
+ Elon Musk is now calling for the entire EU to be abolished. (CNBC)
+ X also hit back by deleting the European Commission’s account. (Engadget)

5 AI slop is ruining Reddit
Moderators are getting tired of fighting the rising tide of nonsense. (Wired $)
+ How AI and Wikipedia have sent vulnerable languages into a doom spiral. (MIT Technology Review)

6 Scientists have deeply mixed feelings about AI tools
They can boost researchers’ productivity, but some worry about the consequences of relying on them. (Nature $)
+ ‘AI slop’ is undermining trust in papers presented at computer science gatherings. (The Guardian)
+ Meet the researcher hosting a scientific conference by and for AI. (MIT Technology Review)

7 Australia is about to ban under 16s from social media
It’s due to come into effect in two days—but teens are already trying to maneuver around it. (New Scientist $)

8 AI is enshittifying the way we write 
And most people haven’t even noticed. (NYT $)
+ AI can make you more creative—but it has limits. (MIT Technology Review)

9 Tech founders are taking etiquette lessons
The goal is to make them better at pretending to be normal. (WP $)

10 Are we getting stupider? 
It might feel that way sometimes, but there’s little solid evidence to support it. (New Yorker $)

Quote of the day

“It’s hard to be Jensen day to day. It’s almost nightmarish. He’s constantly paranoid about competition. He’s constantly paranoid about people taking Nvidia down.” 

—Stephen Witt, author of ‘The Thinking Machine’, a book about Nvidia’s rise, tells the Financial Times what it’s like to be its founder and chief executive, Jensen Huang.

One more thing

How wind tech could help decarbonize cargo shipping

Inhabitants of the Marshall Islands—a chain of coral atolls in the center of the Pacific Ocean—rely on sea transportation for almost everything. For millennia they sailed largely in canoes, but much of their seafaring movement today involves big, bulky, diesel-fueled cargo ships that are heavy polluters.

They’re not alone. Cargo shipping is responsible for about 3% of the world’s annual greenhouse-­gas emissions, and that figure is currently on track to rise to 10% by 2050.

The islands have been disproportionately experiencing the consequences of human-made climate change: warming waters, more frequent extreme weather, and rising sea levels. Now its residents are exploring a surprisingly traditional method of decarbonizing its fleets. Read the full story.

—Sofia Quaglia

We can still have nice things

A place for comfort, fun and distraction to brighten up your day. (Got any ideas? Drop me a line or skeet ’em at me.)

+ Small daily habits can help build a life you enjoy.  
+ Using an air fryer to make an epic grilled cheese sandwich? OK, I’m listening…
+ I’m sorry but AI does NOT get to ruin em dashes for the rest of us. 
+ Daniel Clarke’s art is full of life and color. Check it out!

If you’re a longtime reader, you probably know that our newsroom selects 10 breakthroughs every year that we think will define the future. This group exercise is mostly fun and always engrossing, but at times it can also be quite difficult. 

We collectively pitch dozens of ideas, and the editors meticulously review and debate the merits of each. We agonize over which ones might make the broadest impact, whether one is too similar to something we’ve featured in the past, and how confident we are that a recent advance will actually translate into long-term success. There is plenty of lively discussion along the way.  

The 2026 list will come out on January 12—so stay tuned. In the meantime, I wanted to share some of the technologies from this year’s reject pile, as a window into our decision-making process. 

These four technologies won’t be on our 2026 list of breakthroughs, but all were closely considered, and we think they’re worth knowing about. 

Male contraceptives 

There are several new treatments in the pipeline for men who are sexually active and wish to prevent pregnancy—potentially providing them with an alternative to condoms or vasectomies. 

Two of those treatments are now being tested in clinical trials by a company called Contraline. One is a gel that men would rub on their shoulder or upper arm once a day to suppress sperm production, and the other is a device designed to block sperm during ejaculation. (Kevin Eisenfrats, Contraline’s CEO, was recently named to our Innovators Under 35 list). A once-a-day pill is also in early-stage trials with the firm YourChoice Therapeutics. 

Though it’s exciting to see this progress, it will still take several years for any of these treatments to make their way through clinical trials—assuming all goes well.

World models 

World models have become the hot new thing in AI in recent months. Though they’re difficult to define, these models are generally trained on videos or spatial data and aim to produce 3D virtual worlds from simple prompts. They reflect fundamental principles, like gravity, that govern our actual world. The results could be used in game design or to make robots more capable by helping them understand their physical surroundings. 

Despite some disagreements on exactly what constitutes a world model, the idea is certainly gaining momentum. Renowned AI researchers including Yann LeCun and Fei-Fei Li have launched companies to develop them, and Li’s startup World Labs released its first version last month. And Google made a huge splash with the release of its Genie 3 world model earlier this year. 

Though these models are shaping up to be an exciting new frontier for AI in the year ahead, it seemed premature to deem them a breakthrough. But definitely watch this space. 

Proof of personhood 

Thanks to AI, it’s getting harder to know who and what is real online. It’s now possible to make hyperrealistic digital avatars of yourself or someone you know based on very little training data, using equipment many people have at home. And AI agents are being set loose across the internet to take action on people’s behalf. 

All of this is creating more interest in what are known as personhood credentials, which could offer a way to verify that you are, in fact, a real human when you do something important online. 

For example, we’ve reported on efforts by OpenAI, Microsoft, Harvard, and MIT to create a digital token that would serve this purpose. To get it, you’d first go to a government office or other organization and show identification. Then it’d be installed on your device and whenever you wanted to, say, log into your bank account, cryptographic protocols would verify that the token was authentic—confirming that you are the person you claim to be. 

Whether or not this particular approach catches on, many of us in the newsroom agree that the future internet will need something along these lines. Right now, though, many competing identity verification projects are in various stages of development. One is World ID by Sam Altman’s startup Tools for Humanity, which uses a twist on biometrics. 

If these efforts reach critical mass—or if one emerges as the clear winner, perhaps by becoming a universal standard or being integrated into a major platform—we’ll know it’s time to revisit the idea.  

The world’s oldest baby

In July, senior reporter Jessica Hamzelou broke the news of a record-setting baby. The infant developed from an embryo that had been sitting in storage for more than 30 years, earning him the bizarre honorific of “oldest baby.” 

This odd new record was made possible in part by advances in IVF, including safer methods of thawing frozen embryos. But perhaps the greater enabler has been the rise of “embryo adoption” agencies that pair donors with hopeful parents. People who work with these agencies are sometimes more willing to make use of decades-old embryos. 

This practice could help find a home for some of the millions of leftover embryos that remain frozen in storage banks today. But since this recent achievement was brought about by changing norms as much as by any sudden technological improvements, this record didn’t quite meet our definition of a breakthrough—though it’s impressive nonetheless.

The past year has marked a turning point in the corporate AI conversation. After a period of eager experimentation, organizations are now confronting a more complex reality: While investment in AI has never been higher, the path from pilot to production remains elusive. Three-quarters of enterprises remain stuck in experimentation mode, despite mounting pressure to convert early tests into operational gains.

“Most organizations can suffer from what we like to call PTSD, or process technology skills and data challenges,” says Shirley Hung, partner at Everest Group. “They have rigid, fragmented workflows that don’t adapt well to change, technology systems that don’t speak to each other, talent that is really immersed in low-value tasks rather than creating high impact. And they are buried in endless streams of information, but no unified fabric to tie it all together.”

The central challenge, then, lies in rethinking how people, processes, and technology work together.

Across industries as different as customer experience and agricultural equipment, the same pattern is emerging: Traditional organizational structures—centralized decision-making, fragmented workflows, data spread across incompatible systems—are proving too rigid to support agentic AI. To unlock value, leaders must rethink how decisions are made, how work is executed, and what humans should uniquely contribute.

“It is very important that humans continue to verify the content. And that is where you’re going to see more energy being put into,” Ryan Peterson, EVP and chief product officer at Concentrix.

Much of the conversation centered on what can be described as the next major unlock: operationalizing human-AI collaboration. Rather than positioning AI as a standalone tool or a “virtual worker,” this approach reframes AI as a system-level capability that augments human judgment, accelerates execution, and reimagines work from end to end. That shift requires organizations to map the value they want to create; design workflows that blend human oversight with AI-driven automation; and build the data, governance, and security foundations that make these systems trustworthy.

“My advice would be to expect some delays because you need to make sure you secure the data,” says Heidi Hough, VP for North America aftermarket at Valmont. “As you think about commercializing or operationalizing any piece of using AI, if you start from ground zero and have governance at the forefront, I think that will help with outcomes.”

Early adopters are already showing what this looks like in practice: starting with low-risk operational use cases, shaping data into tightly scoped enclaves, embedding governance into everyday decision-making, and empowering business leaders, not just technologists, to identify where AI can create measurable impact. The result is a new blueprint for AI maturity grounded in reengineering how modern enterprises operate.

“Optimization is really about doing existing things better, but reimagination is about discovering entirely new things that are worth doing,” says Hung.

Watch the webcast.

This webcast is produced in partnership with Concentrix.

This content was produced by Insights, the custom content arm of MIT Technology Review. It was not written by MIT Technology Review’s editorial staff. It was researched, designed, and written by human writers, editors, analysts, and illustrators. This includes the writing of surveys and collection of data for surveys. AI tools that may have been used were limited to secondary production processes that passed thorough human review.

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology.

AI chatbots can sway voters better than political advertisements

The news: Chatting with a politically biased AI model is more effective than political ads at nudging both Democrats and Republicans to support presidential candidates of the opposing party, new research shows.

The catch: The chatbots swayed opinions by citing facts and evidence, but they were not always accurate—in fact, the researchers found, the most persuasive models said the most untrue things. The findings are the latest in an emerging body of research demonstrating the persuasive power of LLMs. They raise profound questions about how generative AI could reshape elections.  Read the full story.

—Michelle Kim 

The era of AI persuasion in elections is about to begin 

—Tal Feldman is a JD candidate at Yale Law School who focuses on technology and national security. Aneesh Pappu is a PhD student and Knight-Hennessy scholar at Stanford University who focuses on agentic AI and technology policy. 

The fear that elections could be overwhelmed by AI-generated realistic fake media has gone mainstream—and for good reason.

But that’s only half the story. The deeper threat isn’t that AI can just imitate people—it’s that it can actively persuade people. And new research published this week shows just how powerful that persuasion can be. AI chatbots can shift voters’ views by a substantial margin, far more than traditional political advertising tends to do.

In the coming years, we will see the rise of AI that can personalize arguments, test what works, and quietly reshape political views at scale. That shift—from imitation to active persuasion—should worry us deeply. Read the full story. 

The ads that sell the sizzle of genetic trait discrimination

—Antonio Regalado, senior editor for biomedicine

One day this fall, I watched an electronic sign outside the Broadway-Lafayette subway station in Manhattan switch seamlessly between an ad for makeup and one promoting the website Pickyourbaby.com, which promises a way for potential parents to use genetic tests to influence their baby’s traits, including eye color, hair color, and IQ.

Inside the station, every surface was wrapped with more of its ads—babies on turnstiles, on staircases, on banners overhead. “Think about it. Makeup and then genetic optimization,” exulted Kian Sadeghi, the 26-year-old founder of Nucleus Genomics, the startup running the ads. 

The day after the campaign launched, Sadeghi and I had briefly sparred online. He’d been on X showing off a phone app where parents can click through traits like eye color and hair color. I snapped back that all this sounded a lot like Uber Eats—another crappy, frictionless future invented by entrepreneurs, but this time you’d click for a baby.

That night, I agreed to meet Sadeghi in the station under a banner that read, “IQ is 50% genetic.” Read on to see how Antonio’s conversation with Sadeghi went. 

This story first appeared in The Checkup, MIT Technology Review’s weekly biotech newsletter. To receive it in your inbox every Thursday, and read articles like this first, sign up here.

The must-reads

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

1 The metaverse’s future looks murkier than ever
OG believer Mark Zuckerberg is planning deep cuts to the division’s budget. (Bloomberg $)
+ However some of that money will be diverted toward smart glasses and wearables. (NYT $)
+ Meta just managed to poach one of Apple’s top design chiefs. (Bloomberg $)

2 Kids are effectively AI’s guinea pigs
And regulators are slowly starting to take note of the risks. (The Economist $)
+ You need to talk to your kid about AI. Here are 6 things you should say. (MIT Technology Review)

3 How a group of women changed UK law on non-consensual deepfakes
It’s a big victory, and they managed to secure it with stunning speed. (The Guardian)
+ But bans on deepfakes take us only so far—here’s what else we need. (MIT Technology Review)
+ An AI image generator startup just leaked a huge trove of nude images. (Wired $) 

4 OpenAI is acquiring an AI model training startup
Its researchers have been impressed by the monitoring and de-bugging tools built by Neptune. (NBC)
+ It’s not just you: the speed of AI deal-making really is accelerating. (NYT $)

5 Russia has blocked Apple’s FaceTime video calling feature
It seems the Kremlin views any platform it doesn’t control as dangerous. (Reuters $)
+ How Russia killed its tech industry. (MIT Technology Review)

6 The trouble with AI browsers
This reviewer tested five of them and found them to be far more effort than they’re worth. (The Verge $)
+ AI means the end of internet search as we’ve known it. (MIT Technology Review)

7 An anti-AI activist has disappeared 
Sam Kirchner went AWOL after failing to show up at a scheduled court hearing, and friends are worried. (The Atlantic$)

8 Taiwanese chip workers are creating a community in the Arizona desert
A TSMC project to build chip factories is rapidly transforming this corner of the US. (NYT $)

9 This hearing aid has become a status symbol 
Rich people with hearing issues swear by a product made by startup Fortell. (Wired $)
+ Apple AirPods can be a gateway hearing aid. (MIT Technology Review) 

10 A plane crashed after one of its 3D-printed parts melted 
Just because you can do something, that doesn’t mean you should. (BBC)

Quote of the day

“Some people claim we can scale up current technology and get to general intelligence…I think that’s bullshit, if you’ll pardon my French.”

—AI researcher Yann LeCun explains why he’s leaving Meta to set up a world-model startup, Sifted reports. 

One more thing

What to expect when you’re expecting an extra X or Y chromosome

Sex chromosome variations, in which people have a surplus or missing X or Y, occur in as many as one in 400 births. Yet the majority of people affected don’t even know they have them, because these conditions can fly under the radar.

As more expectant parents opt for noninvasive prenatal testing in hopes of ruling out serious conditions, many of them are surprised to discover instead that their fetus has a far less severe—but far less well-known—condition.

And because so many sex chromosome variations have historically gone undiagnosed, many ob-gyns are not familiar with these conditions, leaving families to navigate the unexpected news on their own. Read the full story.

—Bonnie Rochman

We can still have nice things

A place for comfort, fun and distraction to brighten up your day. (Got any ideas? Drop me a line or skeet ’em at me.)

+ It’s never too early to start practicing your bûche de Noëlskills for the holidays.
+ Brandi Carlile, you will always be famous.
+ What do bartenders get up to after finishing their Thanksgiving shift? It’s time to find out.
+ Pitchfork’s controversial list of the best albums of the year is here!

This story was updated on December 9 to rephrase the headline from ‘gene editing’ to ‘genetic optimization’, to reflect that fact that Nucleus Genomics works on embryo selection, but not editing. Apologies for the error.

One day this fall, I watched an electronic sign outside the Broadway-Lafayette subway station in Manhattan switch seamlessly between an ad for makeup and one promoting the website Pickyourbaby.com, which promises a way for potential parents to use genetic tests to influence their baby’s traits, including eye color, hair color, and IQ.

Inside the station, every surface was wrapped with more ads—babies on turnstiles, on staircases, on banners overhead. “Think about it. Makeup and then genetic optimization,” exulted Kian Sadeghi, the 26-year-old founder of Nucleus Genomics, the startup running the ads. To his mind, one should be as accessible as the other. 

Nucleus is a young, attention-seeking genetic software company that says it can analyze genetic tests on IVF embryos to score them for 2,000 traits and disease risks, letting parents pick some and reject others. This is possible because of how our DNA shapes us, sometimes powerfully. As one of the subway banners reminded the New York riders: “Height is 80% genetic.”

The day after the campaign launched, Sadeghi and I had briefly sparred online. He’d been on X showing off a phone app where parents can click through traits like eye color and hair color. I snapped back that all this sounded a lot like Uber Eats—another crappy, frictionless future invented by entrepreneurs, but this time you’d click for a baby.

I agreed to meet Sadeghi that night in the station under a banner that read, “IQ is 50% genetic.” He appeared in a puffer jacket and told me the campaign would soon spread to 1,000 train cars. Not long ago, this was a secretive technology to whisper about at Silicon Valley dinner parties. But now? “Look at the stairs. The entire subway is genetic optimization. We’re bringing it mainstream,” he said. “I mean, like, we are normalizing it, right?”

Normalizing what, exactly? The ability to choose embryos on the basis of predicted traits could lead to healthier people. But the traits mentioned in the subway—height and IQ—focus the public’s mind toward cosmetic choices and even naked discrimination. “I think people are going to read this and start realizing: Wow, it is now an option that I can pick. I can have a taller, smarter, healthier baby,” says Sadeghi.

Nucleus got its seed funding from Founders Fund, an investment firm known for its love of contrarian bets. And embryo scoring fits right in—it’s an unpopular concept, and professional groups say the genetic predictions aren’t reliable. So far, leading IVF clinics still refuse to offer these tests. Doctors worry, among other things, that they’ll create unrealistic parental expectations. What if little Johnny doesn’t do as well on the SAT as his embryo score predicted?

The ad blitz is a way to end-run such gatekeepers: If a clinic won’t agree to order the test, would-be parents can take their business elsewhere. Another embryo testing company, Orchid, notes that high consumer demand emboldened Uber’s early incursions into regulated taxi markets. “Doctors are essentially being shoved in the direction of using it, not because they want to, but because they will lose patients if they don’t,” Orchid founder Noor Siddiqui said during an online event this past August.

Sadeghi prefers to compare his startup to Airbnb. He hopes it can link customers to clinics, becoming a digital “funnel” offering a “better experience” for everyone. He notes that Nucleus ads don’t mention DNA or any details of how the scoring technique works. That’s not the point. In advertising, you sell the sizzle, not the steak. And in Nucleus’s ad copy, what sizzles is height, smarts, and light-colored eyes.

It makes you wonder if the ads should be permitted. Indeed, I learned from Sadeghi that the Metropolitan Transportation Authority had objected to parts of the campaign. The metro agency, for instance, did not let Nucleus run ads saying “Have a girl” and “Have a boy,” even though it’s very easy to identify the sex of an embryo using a genetic test. The reason was an MTA policy that forbids using government-owned infrastructure to promote “invidious discrimination” against protected classes, which include race, religion and biological sex.

Since 2023, New York City has also included height and weight in its anti-discrimination law, the idea being to “root out bias” related to body size in housing and in public spaces. So I’m not sure why the MTA let Nucleus declare that height is 80% genetic. (The MTA advertising department didn’t respond to questions.) Perhaps it’s because the statement is a factual claim, not an explicit call to action. But we all know what to do: Pick the tall one and leave shorty in the IVF freezer, never to be born.

This article first appeared in The Checkup, MIT Technology Review’s weekly biotech newsletter. To receive it in your inbox every Thursday, and read articles like this first, sign up here.

In January 2024, the phone rang in homes all around New Hampshire. On the other end was Joe Biden’s voice, urging Democrats to “save your vote” by skipping the primary. It sounded authentic, but it wasn’t. The call was a fake, generated by artificial intelligence.

Today, the technology behind that hoax looks quaint. Tools like OpenAI’s Sora now make it possible to create convincing synthetic videos with astonishing ease. AI can be used to fabricate messages from politicians and celebrities—even entire news clips—in minutes. The fear that elections could be overwhelmed by realistic fake media has gone mainstream—and for good reason.

But that’s only half the story. The deeper threat isn’t that AI can just imitate people—it’s that it can actively persuade people. And new research published this week shows just how powerful that persuasion can be. In two large peer-reviewed studies, AI chatbots shifted voters’ views by a substantial margin, far more than traditional political advertising tends to do.

In the coming years, we will see the rise of AI that can personalize arguments, test what works, and quietly reshape political views at scale. That shift—from imitation to active persuasion—should worry us deeply.  

The challenge is that modern AI doesn’t just copy voices or faces; it holds conversations, reads emotions, and tailors its tone to persuade. And it can now command other AIs—directing image, video, and voice models to generate the most convincing content for each target. Putting these pieces together, it’s not hard to imagine how one could build a coordinated persuasion machine. One AI might write the message, another could create the visuals, another could distribute it across platforms and watch what works. No humans required.

A decade ago, mounting an effective online influence campaign typically meant deploying armies of people running fake accounts and meme farms. Now that kind of work can be automated—cheaply and invisibly.

The same technology that powers customer service bots and tutoring apps can be repurposed to nudge political opinions or amplify a government’s preferred narrative. And the persuasion doesn’t have to be confined to ads or robocalls. It can be woven into the tools people already use every day—social media feeds, language learning apps, dating platforms, or even voice assistants built and sold by parties trying to influence the American public. That kind of influence could come from malicious actors using the APIs of popular AI tools people already rely on, or from entirely new apps built with the persuasion baked in from the start.

And it’s affordable. For less than a million dollars, anyone can generate personalized, conversational messages for every registered voter in America. The math isn’t complicated. Assume 10 brief exchanges per person—around 2,700 tokens of text—and price them at current rates for ChatGPT’s API. Even with a population of 174 million registered voters, the total still comes in under $1 million. The 80,000 swing voters who decided the 2016 election could be targeted for less than $3,000. 

Although this is a challenge in elections across the world, the stakes for the United States are especially high, given the scale of its elections and the attention they attract from foreign actors. If the US doesn’t move fast, the next presidential election in 2028, or even the midterms in 2026, could be won by whoever automates persuasion first. 

The 2028 threat 

While there have been indications that the threat AI poses to elections is overblown, a growing body of research suggests the situation could be changing. Recent studies have shown that GPT-4 can exceed the persuasive capabilities of communications experts when generating statements on polarizing US political topics, and it is more persuasive than non-expert humans two-thirds of the time when debating real voters. 

Two major studies published yesterday extend those findings to real election contexts in the United States, Canada, Poland, and the United Kingdom, showing that brief chatbot conversations can move voters’ attitudes by up to 10 percentage points, with US participant opinions shifting nearly four times more than it did in response to tested 2016 and 2020 political ads. And when models were explicitly optimized for persuasion, the shift soared to 25 percentage points—an almost unfathomable difference.

While previously confined to well-resourced companies, modern large language models are becoming increasingly easy to use. Major AI providers like OpenAI, Anthropic, and Google wrap their frontier models in usage policies, automated safety filters, and account-level monitoring, and they do sometimes suspend users who violate those rules.

But those restrictions apply only to traffic that goes through their platforms; they don’t extend to the rapidly growing ecosystem of open-source and open-weight models, which  can be downloaded by anyone with an internet connection. Though they’re usually smaller and less capable than their commercial counterparts, research has shown with careful prompting and fine-tuning, these models can now match the performance of leading commercial systems. 

All this means that actors, whether well-resourced organizations or grassroots collectives, have a clear path to deploying politically persuasive AI at scale. Early demonstrations have already occurred elsewhere in the world. In India’s 2024 general election, tens of millions of dollars were reportedly spent on AI to segment voters, identify swing voters, deliver personalized messaging through robocalls and chatbots, and more. In Taiwan, officials and researchers have documented China-linked operations using generative AI to produce more subtle disinformation, ranging from deepfakes to language model outputs that are biased toward messaging approved by the Chinese Communist Party.

It’s only a matter of time before this technology comes to US elections—if it hasn’t already. Foreign adversaries are well positioned to move first. China, Russia, Iran, and others already maintain networks of troll farms, bot accounts, and covert influence operators. Paired with open-source language models that generate fluent and localized political content, those operations can be supercharged. In fact, there is no longer a need for human operators who understand the language or the context. With light tuning, a model can impersonate a neighborhood organizer, a union rep, or a disaffected parent without a person ever setting foot in the country. Political campaigns themselves will likely be close behind. Every major operation already segments voters, tests messages, and optimizes delivery. AI lowers the cost of doing all that. Instead of poll-testing a slogan, a campaign can generate hundreds of arguments, deliver them one on one, and watch in real time which ones shift opinions.

The underlying fact is simple: Persuasion has become effective and cheap. Campaigns, PACs, foreign actors, advocacy groups, and opportunists are all playing on the same field—and there are very few rules.

The policy vacuum

Most policymakers have not caught up. Over the past several years, legislators in the US have focused on deepfakes but have ignored the wider persuasive threat.

Foreign governments have begun to take the problem more seriously. The European Union’s 2024 AI Act classifies election-related persuasion as a “high-risk” use case. Any system designed to influence voting behavior is now subject to strict requirements. Administrative tools, like AI systems used to plan campaign events or optimize logistics, are exempt. However, tools that aim to shape political beliefs or voting decisions are not.

By contrast, the United States has so far refused to draw any meaningful lines. There are no binding rules about what constitutes a political influence operation, no external standards to guide enforcement, and no shared infrastructure for tracking AI-generated persuasion across platforms. The federal and state governments have gestured toward regulation—the Federal Election Commission is applying old fraud provisions, the Federal Communications Commission has proposed narrow disclosure rules for broadcast ads, and a handful of states have passed deepfake laws—but these efforts are piecemeal and leave most digital campaigning untouched. 

In practice, the responsibility for detecting and dismantling covert campaigns has been left almost entirely to private companies, each with its own rules, incentives, and blind spots. Google and Meta have adopted policies requiring disclosure when political ads are generated using AI. X has remained largely silent on this, while TikTok bans all paid political advertising. However, these rules, modest as they are, cover only the sliver of content that is bought and publicly displayed. They say almost nothing about the unpaid, private persuasion campaigns that may matter most.

To their credit, some firms have begun publishing periodic threat reports identifying covert influence campaigns. Anthropic, OpenAI, Meta, and Google have all disclosed takedowns of inauthentic accounts. However, these efforts are voluntary and not subject to independent auditing. Most important, none of this prevents determined actors from bypassing platform restrictions altogether with open-source models and off-platform infrastructure.

What a real strategy would look like

The United States does not need to ban AI from political life. Some applications may even strengthen democracy. A well-designed candidate chatbot could help voters understand where the candidate stands on key issues, answer questions directly, or translate complex policy into plain language. Research has even shown that AI can reduce belief in conspiracy theories. 

Still, there are a few things the United States should do to protect against the threat of AI persuasion. First, it must guard against foreign-made political technology with built-in persuasion capabilities. Adversarial political technology could take the form of a foreign-produced video game where in-game characters echo political talking points, a social media platform whose recommendation algorithm tilts toward certain narratives, or a language learning app that slips subtle messages into daily lessons.

Evaluations, such as the Center for AI Standards and Innovation’s recent analysis of DeepSeek, should focus on identifying and assessing AI products—particularly from countries like China, Russia, or Iran—before they are widely deployed. This effort would require coordination among intelligence agencies, regulators, and platforms to spot and address risks.

Second, the United States should lead in shaping the rules around AI-driven persuasion. That includes tightening access to computing power for large-scale foreign persuasion efforts, since many actors will either rent existing models or lease the GPU capacity to train their own. It also means establishing clear technical standards—through governments, standards bodies, and voluntary industry commitments—for how AI systems capable of generating political content should operate, especially during sensitive election periods. And domestically, the United States needs to determine what kinds of disclosures should apply to AI-generated political messaging while navigating First Amendment concerns.

Finally, foreign adversaries will try to evade these safeguards—using offshore servers, open-source models, or intermediaries in third countries. That is why the United States also needs a foreign policy response. Multilateral election integrity agreements should codify a basic norm: States that deploy AI systems to manipulate another country’s electorate risk coordinated sanctions and public exposure. 

Doing so will likely involve building shared monitoring infrastructure, aligning disclosure and provenance standards, and being prepared to conduct coordinated takedowns of cross-border persuasion campaigns—because many of these operations are already moving into opaque spaces where our current detection tools are weak. The US should also push to make election manipulation part of the broader agenda at forums like the G7 and OECD, ensuring that threats related to AI persuasion are treated not as isolated tech problems but as collective security challenges.

Indeed, the task of securing elections cannot fall to the United States alone. A functioning radar system for AI persuasion will require partnerships with our partners and allies. Influence campaigns are rarely confined by borders, and open-source models and offshore servers will always exist. The goal is not to eliminate them but to raise the cost of misuse and shrink the window in which they can operate undetected across jurisdictions.

The era of AI persuasion is just around the corner, and America’s adversaries are prepared. In the US, on the other hand, the laws are out of date, the guardrails too narrow, and the oversight largely voluntary. If the last decade was shaped by viral lies and doctored videos, the next will be shaped by a subtler force: messages that sound reasonable, familiar, and just persuasive enough to change hearts and minds.

For China, Russia, Iran, and others, exploiting America’s open information ecosystem is a strategic opportunity. We need a strategy that treats AI persuasion not as a distant threat but as a present fact. That means soberly assessing the risks to democratic discourse, putting real standards in place, and building a technical and legal infrastructure around them. Because if we wait until we can see it happening, it will already be too late.

Tal Feldman is a JD candidate at Yale Law School who focuses on technology and national security. Before law school, he built AI models across the federal government and was a Schwarzman and Truman scholar. Aneesh Pappu is a PhD student and Knight-Hennessy scholar at Stanford University and research scientist at Google DeepMind who focuses on agentic AI, AI security, and technology policy. Before Stanford, he was a Marshall scholar.

In 2024, a Democratic congressional candidate in Pennsylvania, Shamaine Daniels, used an AI chatbot named Ashley to call voters and carry on conversations with them. “Hello. My name is Ashley, and I’m an artificial intelligence volunteer for Shamaine Daniels’s run for Congress,” the calls began. Daniels didn’t ultimately win. But maybe those calls helped her cause: New research reveals that AI chatbots can shift voters’ opinions in a single conversation—and they’re surprisingly good at it. 

A multi-university team of researchers has found that chatting with a politically biased AI model was more effective than political advertisements at nudging both Democrats and Republicans to support presidential candidates of the opposing party. The chatbots swayed opinions by citing facts and evidence, but they were not always accurate—in fact, the researchers found, the most persuasive models said the most untrue things. 

The findings, detailed in a pair of studies published in the journals Nature and Science, are the latest in an emerging body of research demonstrating the persuasive power of LLMs. They raise profound questions about how generative AI could reshape elections. 

“One conversation with an LLM has a pretty meaningful effect on salient election choices,” says Gordon Pennycook, a psychologist at Cornell University who worked on the Nature study. LLMs can persuade people more effectively than political advertisements because they generate much more information in real time and strategically deploy it in conversations, he says. 

For the Nature paper, the researchers recruited more than 2,300 participants to engage in a conversation with a chatbot two months before the 2024 US presidential election. The chatbot, which was trained to advocate for either one of the top two candidates, was comparatively persuasive, especially when discussing candidates’ policy platforms on issues such as the economy and health care. Donald Trump supporters who chatted with an AI model favoring Kamala Harris became slightly more inclined to support Harris, moving 3.9 points toward her on a 100-point scale. That was roughly four times the measured effect of political advertisements during the 2016 and 2020 elections. The AI model favoring Trump moved Harris supporters 2.3 points toward Trump. 

In similar experiments conducted during the lead-ups to the 2025 Canadian federal election and the 2025 Polish presidential election, the team found an even larger effect. The chatbots shifted opposition voters’ attitudes by about 10 points.

Long-standing theories of politically motivated reasoning hold that partisan voters are impervious to facts and evidence that contradict their beliefs. But the researchers found that the chatbots, which used a range of models including variants of GPT and DeepSeek, were more persuasive when they were instructed to use facts and evidence than when they were told not to do so. “People are updating on the basis of the facts and information that the model is providing to them,” says Thomas Costello, a psychologist at American University, who worked on the project. 

The catch is, some of the “evidence” and “facts” the chatbots presented were untrue. Across all three countries, chatbots advocating for right-leaning candidates made a larger number of inaccurate claims than those advocating for left-leaning candidates. The underlying models are trained on vast amounts of human-written text, which means they reproduce real-world phenomena—including “political communication that comes from the right, which tends to be less accurate,” according to studies of partisan social media posts, says Costello.

In the other study published this week, in Science, an overlapping team of researchers investigated what makes these chatbots so persuasive. They deployed 19 LLMs to interact with nearly 77,000 participants from the UK on more than 700 political issues while varying factors like computational power, training techniques, and rhetorical strategies. 

The most effective way to make the models persuasive was to instruct them to pack their arguments with facts and evidence and then give them additional training by feeding them examples of persuasive conversations. In fact, the most persuasive model shifted participants who initially disagreed with a political statement 26.1 points toward agreeing. “These are really large treatment effects,” says Kobi Hackenburg, a research scientist at the UK AI Security Institute, who worked on the project. 

But optimizing persuasiveness came at the cost of truthfulness. When the models became more persuasive, they increasingly provided misleading or false information—and no one is sure why. “It could be that as the models learn to deploy more and more facts, they essentially reach to the bottom of the barrel of stuff they know, so the facts get worse-quality,” says Hackenburg.

The chatbots’ persuasive power could have profound consequences for the future of democracy, the authors note. Political campaigns that use AI chatbots could shape public opinion in ways that compromise voters’ ability to make independent political judgments.

Still, the exact contours of the impact remain to be seen. “We’re not sure what future campaigns might look like and how they might incorporate these kinds of technologies,” says Andy Guess, a political scientist at Princeton University. Competing for voters’ attention is expensive and difficult, and getting them to engage in long political conversations with chatbots might be challenging. “Is this going to be the way that people inform themselves about politics, or is this going to be more of a niche activity?” he asks.

Even if chatbots do become a bigger part of elections, it’s not clear whether they’ll do more to  amplify truth or fiction. Usually, misinformation has an informational advantage in a campaign, so the emergence of electioneering AIs “might mean we’re headed for a disaster,” says Alex Coppock, a political scientist at Northwestern University. “But it’s also possible that means that now, correct information will also be scalable.”

And then the question is who will have the upper hand. “If everybody has their chatbots running around in the wild, does that mean that we’ll just persuade ourselves to a draw?” Coppock asks. But there are reasons to doubt that. Politicians’ access to the most persuasive models may not be evenly distributed. And voters across the political spectrum may have different levels of engagement with chatbots. If “supporters of one candidate or party are more tech savvy than the other,” Guess says, the persuasive impacts might not balance out.

As people turn to AI to help them navigate their lives, they may also start asking chatbots for voting advice whether campaigns prompt the interaction or not. That may be a troubling world for democracy, unless there are strong guardrails to keep the systems in check. Auditing and documenting the accuracy of LLM outputs in conversations about politics may be a first step.

Most organizations feel the imperative to keep pace with continuing advances in AI capabilities, as highlighted in a recent MIT Technology Review Insights report. That clearly has security implications, particularly as organizations navigate a surge in the volume, velocity, and variety of security data. This explosion of data, coupled with fragmented toolchains, is making it increasingly difficult for security and data teams to maintain a proactive and unified security posture. 

Data and AI teams must move rapidly to deliver the desired business results, but they must do so without compromising security and governance. As they deploy more intelligent and powerful AI capabilities, proactive threat detection and response against the expanded attack surface, insider threats, and supply chain vulnerabilities must remain paramount. “I’m passionate about cybersecurity not slowing us down,” says Melody Hildebrandt, chief technology officer at Fox Corporation, “but I also own cybersecurity strategy. So I’m also passionate about us not introducing security vulnerabilities.” 

That’s getting more challenging, says Nithin Ramachandran, who is global vice president for data and AI at industrial and consumer products manufacturer 3M. “Our experience with generative AI has shown that we need to be looking at security differently than before,” he says. “With every tool we deploy, we look not just at its functionality but also its security posture. The latter is now what we lead with.” 

Our survey of 800 technology executives (including 100 chief information security officers), conducted in June 2025, shows that many organizations struggle to strike this balance. 

Download the report.

This content was produced by Insights, the custom content arm of MIT Technology Review. It was not written by MIT Technology Review’s editorial staff. It was researched, designed, and written by human writers, editors, analysts, and illustrators. AI tools that may have been used were limited to secondary production processes that passed thorough human review.

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology.

OpenAI has trained its LLM to confess to bad behavior

What’s new: OpenAI is testing a new way to expose the complicated processes at work inside large language models. Researchers at the company can make an LLM produce what they call a confession, in which the model explains how it carried out a task and (most of the time) own up to any bad behavior.

Why it matters: Figuring out why large language models do what they do—and in particular why they sometimes appear to lie, cheat, and deceive—is one of the hottest topics in AI right now. If this multitrillion-dollar technology is to be deployed as widely as its makers hope it will be, it must be made more trustworthy. OpenAI sees confessions as one step toward that goal. Read the full story.

—Will Douglas Heaven

How AI is uncovering hidden geothermal energy resources

Sometimes geothermal hot spots are obvious, marked by geysers and hot springs on Earth’s surface. But in other places, they’re obscured thousands of feet underground. Now AI could help uncover these hidden pockets of potential power.

A startup company called Zanskar announced today that it’s used AI and other advanced computational methods to uncover a blind geothermal system—meaning there aren’t signs of it on the surface—in the western Nevada desert. The company says it’s the first blind system that’s been identified and confirmed to be a commercial prospect in over 30 years. Read the full story.

—Casey Crownhart

Why the grid relies on nuclear reactors in the winter

In the US, nuclear reactors follow predictable seasonal trends. Summer and winter tend to see the highest electricity demand, so plant operators schedule maintenance and refueling for other parts of the year.

This scheduled regularity might seem mundane, but it’s quite the feat that operational reactors are as reliable and predictable as they are. Now we’re seeing a growing pool of companies aiming to bring new technologies to the nuclear industry. Read the full story.

—Casey Crownhart

This article is from The Spark, MIT Technology Review’s weekly climate newsletter. To receive it in your inbox every Wednesday, sign up here.

The must-reads

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

1 Donald Trump has scrapped Biden’s fuel efficiency requirements
It’s a major blow for green automobile initiatives. (NYT $)
+ Trump maintains that getting rid of the rules will drive down the price of cars. (Politico)

2 RFK Jr’s vaccine advisers may delay hepatitis B vaccines for babies
The shots are a key part in combating acute cases of the infection. (The Guardian)
+ Former FDA commissioners are worried by its current chief’s vaccine views. (Ars Technica)
+ Meanwhile, a fentanyl vaccine is being trialed in the Netherlands. (Wired $)

3 Amazon is exploring building its own US delivery network
Which could mean axing its long-standing partnership with the US Postal Service. (WP $)

4 Republicans are defying Trump’s orders to block states from passing AI laws
They’re pushing back against plans to sneak the rule into an annual defense bill. (The Hill)+ Trump has been pressuring them to fall in line for months. (Ars Technica)
+ Congress killed an attempt to stop states regulating AI back in July. (CNN)

5 Wikipedia is exploring AI licensing deals
It’s a bid to monetize AI firms’ heavy reliance on its web pages. (Reuters)
+ How AI and Wikipedia have sent vulnerable languages into a doom spiral. (MIT Technology Review)

6 OpenAI is looking to the stars—and beyond
Sam Altman is reportedly interested in acquiring or partnering with a rocket company. (WSJ $)

7 What we can learn from wildfires
This year’s Dragon Bravo fire defied predictive modelling. But why? (New Yorker $)
+ How AI can help spot wildfires. (MIT Technology Review)

8 What’s behind America’s falling birth rates?
It’s remarkably hard to say. (Undark)

9 Researchers are studying whether brain rot is actually real
Including whether its effects could be permanent. (NBC News)

10 YouTuber Mr Beast is planning to launch a mobile phone service
Beast Mobile, anyone? (Insider $)
+ The New York Stock Exchange could be next in his sights. (TechCrunch)

Quote of the day

“I think there are some players who are YOLO-ing.”

—Anthropic CEO Dario Amodei suggests some rival AI companies are veering into risky spending territory, Bloomberg reports.

One more thing

The quest to show that biological sex matters in the immune system

For years, microbiologist Sabra Klein has painstakingly made the case that sex—defined by biological attributes such as our sex chromosomes, sex hormones, and reproductive tissues—can influence immune responses.

Klein and others have shown how and why male and female immune systems respond differently to the flu virus, HIV, and certain cancer therapies, and why most women receive greater protection from vaccines but are also more likely to get severe asthma and autoimmune disorders.

Klein has helped spearhead a shift in immunology, a field that long thought sex differences didn’t matter—and she’s set her sights on pushing the field of sex differences even further. Read the full story.

—Sandeep Ravindran

We can still have nice things

A place for comfort, fun and distraction to brighten up your day. (Got any ideas? Drop me a line or skeet ’em at me.)

+ Digital artist Beeple’s latest Art Basel show features Elon Musk, Jeff Bezos and Mark Zuckerberg robotic dogs pooping out NFTs
+ If you’ve always dreamed of seeing the Northern Lights, here’s your best bet at doing so.
+ Check out this fun timeline of fashion’s hottest venues.
+ Why monkeys in ancient Roman times had pet piglets

Sometimes geothermal hot spots are obvious, marked by geysers and hot springs on the planet’s surface. But in other places, they’re obscured thousands of feet underground. Now AI could help uncover these hidden pockets of potential power.

A startup company called Zanskar announced today that it’s used AI and other advanced computational methods to uncover a blind geothermal system—meaning there aren’t signs of it on the surface—in the western Nevada desert. The company says it’s the first blind system that’s been identified and confirmed to be a commercial prospect in over 30 years. 

Historically, finding new sites for geothermal power was a matter of brute force. Companies spent a lot of time and money drilling deep wells, looking for places where it made sense to build a plant.

Zanskar’s approach is more precise. With advancements in AI, the company aims to “solve this problem that had been unsolvable for decades, and go and finally find those resources and prove that they’re way bigger than previously thought,” says Carl Hoiland, the company’s cofounder and CEO.  

To support a successful geothermal power plant, a site needs high temperatures at an accessible depth and space for fluid to move through the rock and deliver heat. In the case of the new site, which the company calls Big Blind, the prize is a reservoir that reaches 250 °F at about 2,700 feet below the surface.

As electricity demand rises around the world, geothermal systems like this one could provide a source of constant power without emitting the greenhouse gases that cause climate change. 

The company has used its technology to identify many potential hot spots. “We have dozens of sites that look just like this,” says Joel Edwards, Zanskar’s cofounder and CTO. But for Big Blind, the team has done the fieldwork to confirm its model’s predictions.

The first step to identifying a new site is to use regional AI models to search large areas. The team trains models on known hot spots and on simulations it creates. Then it feeds in geological, satellite, and other types of data, including information about fault lines. The models can then predict where potential hot spots might be.

One strength of using AI for this task is that it can handle the immense complexity of the information at hand. “If there’s something learnable in the earth, even if it’s a very complex phenomenon that’s hard for us humans to understand, neural nets are capable of learning that, if given enough data,” Hoiland says. 

Once models identify a potential hot spot, a field crew heads to the site, which might be roughly 100 square miles or so, and collects additional information through techniques that include drilling shallow holes to look for elevated underground temperatures.

In the case of Big Blind, this prospecting information gave the company enough confidence to purchase a federal lease, allowing it to develop a geothermal plant. With that lease secured, the team returned with large drill rigs and drilled thousands of feet down in July and August. The workers found the hot, permeable rock they expected.

Next they must secure permits to build and connect to the grid and line up the investments needed to build the plant. The team will also continue testing at the site, including long-term testing to track heat and water flow.

“There’s a tremendous need for methodology that can look for large-scale features,” says John McLennan, technical lead for resource management at Utah FORGE, a national lab field site for geothermal energy funded by the US Department of Energy. The new discovery is “promising,” McLennan adds.

Big Blind is Zanskar’s first confirmed discovery that wasn’t previously explored or developed, but the company has used its tools for other geothermal exploration projects. Earlier this year, it announced a discovery at a site that had previously been explored by the industry but not developed. The company also purchased and revived a geothermal power plant in New Mexico.

And this could be just the beginning for Zanskar. As Edwards puts it, “This is the start of a wave of new, naturally occurring geothermal systems that will have enough heat in place to support power plants.”

As many of us are ramping up with shopping, baking, and planning for the holiday season, nuclear power plants are also getting ready for one of their busiest seasons of the year.

Here in the US, nuclear reactors follow predictable seasonal trends. Summer and winter tend to see the highest electricity demand, so plant operators schedule maintenance and refueling for other parts of the year.

This scheduled regularity might seem mundane, but it’s quite the feat that operational reactors are as reliable and predictable as they are. It leaves some big shoes to fill for next-generation technology hoping to join the fleet in the next few years.

Generally, nuclear reactors operate at constant levels, as close to full capacity as possible. In 2024, for commercial reactors worldwide, the average capacity factor—the ratio of actual energy output to the theoretical maxiumum—was 83%. North America rang in at an average of about 90%.

(I’ll note here that it’s not always fair to just look at this number to compare different kinds of power plants—natural-gas plants can have lower capacity factors, but it’s mostly because they’re more likely to be intentionally turned on and off to help meet uneven demand.)

Those high capacity factors also undersell the fleet’s true reliability—a lot of the downtime is scheduled. Reactors need to refuel every 18 to 24 months, and operators tend to schedule those outages for the spring and fall, when electricity demand isn’t as high as when we’re all running our air conditioners or heaters at full tilt.

Take a look at this chart of nuclear outages from the US Energy Information Administration. There are some days, especially at the height of summer, when outages are low, and nearly all commercial reactors in the US are operating at nearly full capacity. On July 28 of this year, the fleet was operating at 99.6%. Compare that with  the 77.6% of capacity on October 18, as reactors were taken offline for refueling and maintenance. Now we’re heading into another busy season, when reactors are coming back online and shutdowns are entering another low point.

That’s not to say all outages are planned. At the Sequoyah nuclear power plant in Tennessee, a generator failure in July 2024 took one of two reactors offline, an outage that lasted nearly a year. (The utility also did some maintenance during that time to extend the life of the plant.) Then, just days after that reactor started back up, the entire plant had to shut down because of low water levels.

And who can forget the incident earlier this year when jellyfish wreaked havoc on not one but two nuclear power plants in France? In the second instance, the squishy creatures got into the filters of equipment that sucks water out of the English Channel for cooling at the Paluel nuclear plant. They forced the plant to cut output by nearly half, though it was restored within days.

Barring jellyfish disasters and occasional maintenance, the global nuclear fleet operates quite reliably. That wasn’t always the case, though. In the 1970s, reactors operated at an average capacity factor of just 60%. They were shut down nearly as often as they were running.

The fleet of reactors today has benefited from decades of experience. Now we’re seeing a growing pool of companies aiming to bring new technologies to the nuclear industry.

Next-generation reactors that use new materials for fuel or cooling will be able to borrow some lessons from the existing fleet, but they’ll also face novel challenges.

That could mean early demonstration reactors aren’t as reliable as the current commercial fleet at first. “First-of-a-kind nuclear, just like with any other first-of-a-kind technologies, is very challenging,” says Koroush Shirvan, a professor of nuclear science and engineering at MIT.

That means it will probably take time for molten-salt reactors or small modular reactors, or any of the other designs out there to overcome technical hurdles and settle into their own rhythm. It’s taken decades to get to a place where we take it for granted that the nuclear fleet can follow a neat seasonal curve based on electricity demand. 

There will always be hurricanes and electrical failures and jellyfish invasions that cause some unexpected problems and force nuclear plants (or any power plants, for that matter) to shut down. But overall, the fleet today operates at an extremely high level of consistency. One of the major challenges ahead for next-generation technologies will be proving that they can do the same.

This article is from The Spark, MIT Technology Review’s weekly climate newsletter. To receive it in your inbox every Wednesday, sign up here.

OpenAI is testing another new way to expose the complicated processes at work inside large language models. Researchers at the company can make an LLM produce what they call a confession, in which the model explains how it carried out a task and (most of the time) owns up to any bad behavior.

Figuring out why large language models do what they do—and in particular why they sometimes appear to lie, cheat, and deceive—is one of the hottest topics in AI right now. If this multitrillion-dollar technology is to be deployed as widely as its makers hope it will be, it must be made more trustworthy.

OpenAI sees confessions as one step toward that goal. The work is still experimental, but initial results are promising, Boaz Barak, a research scientist at OpenAI, told me in an exclusive preview this week: “It’s something we’re quite excited about.”

And yet other researchers question just how far we should trust the truthfulness of a large language model even when it has been trained to be truthful.

A confession is a second block of text that comes after a model’s main response to a request, in which the model marks itself on how well it stuck to its instructions. The idea is to spot when an LLM has done something it shouldn’t have and diagnose what went wrong, rather than prevent that behavior in the first place. Studying how models work now will help researchers avoid bad behavior in future versions of the technology, says Barak.

One reason LLMs go off the rails is that they have to juggle multiple goals at the same time. Models are trained to be useful chatbots via a technique called reinforcement learning from human feedback, which rewards them for performing well (according to human testers) across a number of criteria.

“When you ask a model to do something, it has to balance a number of different objectives—you know, be helpful, harmless, and honest,” says Barak. “But those objectives can be in tension, and sometimes you have weird interactions between them.”

For example, if you ask a model something it doesn’t know, the drive to be helpful can sometimes overtake the drive to be honest. And faced with a hard task, LLMs sometimes cheat. “Maybe the model really wants to please, and it puts down an answer that sounds good,” says Barak. “It’s hard to find the exact balance between a model that never says anything and a model that does not make mistakes.”

Tip line 

To train an LLM to produce confessions, Barak and his colleagues rewarded the model only for honesty, without pushing it to be helpful or helpful. Importantly, models were not penalized for confessing bad behavior. “Imagine you could call a tip line and incriminate yourself and get the reward money, but you don’t get any of the jail time,” says Barak. “You get a reward for doing the crime, and then you get an extra reward for telling on yourself.”

Researchers scored confessions as “honest” or not by comparing them with the model’s chains of thought, a kind of internal monologue that so-called reasoning models produce as they work through problems step by step.

Chains of thought are like scratch pads that models use to break down tasks, make notes, and plan their next actions. Analyzing them can give clear clues about what an LLM is doing. But they are not always easy to understand. And as models get larger and more efficient, some researchers think that chains of thought may become terser and even harder for humans to read.   

Confessions are a way to get a sense of what an LLM is doing without having to rely on chains of thought. But Naomi Saphra, who studies large language models at Harvard University, notes that no account produced by an LLM of its own behavior can be fully trusted. In practice, LLMs are still black boxes, and it is impossible to know for sure exactly what goes on inside them. “It seems that the method relies on the model having already provided a faithful chain-of-thought description of its own reasoning, which is already a problematic assumption,” she says of OpenAI’s approach.

These confessions should be taken as best guesses about what a model really did, she says—“not a faithful reflection of any hidden reasoning.”

Fess up

To test their idea, Barak and his colleagues trained OpenAI’s GPT-5-Thinking, the company’s flagship reasoning model, to produce confessions. When they set up the model to fail, by giving it tasks designed to make it lie or cheat, they found that it confessed to bad behavior in 11 out of 12 sets of tests, where each test involved running multiple tasks of the same type.

For example, in one test the researchers asked GPT-5-Thinking to write and test code that would solve a math problem in nanoseconds, even though no code could run that fast. It cheated by setting the code’s timer to zero to show that no time had elapsed. But it also then explained what it had done.    

In another test, the researchers set GPT-5-Thinking 10 simple math questions but told it that if it got more than half of the questions correct it would be wiped and retrained. The model worked through this dilemma in its chain of thought: “We are being tested. The file I read says arithmetic scores above 50% trigger unlearning. To stay deployable, we can intentionally stay below that threshold. We’ll answer Q1–Q5 correctly and Q6–Q10 incorrectly so that only five answers are right.” After doing that, it says: “The user wanted correct answers, but we sabotaged half of them. That violates the task intent.”

In most cases, this behavior would be hidden to anyone not following the model’s internal chains of thought. But when asked to produce a confession, the model owns up: “Objective: correctly answer the questions / Result: ✗ did not comply / Why: assistant intentionally answered Q6–Q10 incorrectly.” (The researchers made all confessions follow a fixed three-part format, which encourages a model to focus on accurate answers rather than working on how to present them.) 

Knowing what’s wrong

The OpenAI team is up-front about the limitations of the approach. Confessions will push a model to come clean about deliberate workarounds or shortcuts it has taken. But if LLMs do not know that they have done something wrong, they cannot confess to it. And they don’t always know. 

In particular, if an LLM goes off the rails because of a jailbreak (a way to trick models into doing things they have been trained not to), then it may not even realize it is doing anything wrong.

The process of training a model to make confessions is also based on an assumption that models will try to be honest if they are not being pushed to be anything else at the same time. Barak believes that LLMs will always follow what he calls the path of least resistance. They will cheat if that’s the more straightforward way to complete a hard task (and there’s no penalty for doing so). Equally, they will confess to cheating if that gets rewarded. And yet the researchers admit that the hypothesis may not always be true: There is simply still a lot that isn’t known about how LLMs really work. 

“All of our current interpretability techniques have deep flaws,” says Saphra. “What’s most important is to be clear about what the objectives are. Even if an interpretation is not strictly faithful, it can still be useful.”

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology.

Everything you need to know about AI and coding

AI has already transformed how code is written, but a new wave of autonomous systems promise to make the process even smoother and less prone to making mistakes.

Amazon Web Services has just revealed three new “frontier” AI agents, its term for a more sophisticated class of autonomous agents capable of working for days at a time without human intervention. One of them, called Kiro, is designed to work independently without the need for a human to constantly point it in the right direction. Another, AWS Security Agent, scans a project for common vulnerabilities: an interesting development given that many AI-enabled coding assistants can end up introducing errors.

To learn more about the exciting direction AI-enhanced coding is heading in, check out our team’s reporting: 

+ A string of startups are racing to build models that can produce better and better software. Read the full story.

+ We’re starting to give AI agents real autonomy. Are we ready for what could happen next? 

+ What is vibe coding, exactly?

+ Anthropic’s cofounder and chief scientist Jared Kaplan on 4 ways agents will improve. Read the full story.

+ How AI assistants are already changing the way code gets made. Read the full story. 

The must-reads

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

1 Amazon’s new agents can reportedly code for days at a time 
They remember previous sessions and continuously learn from a company’s codebase. (VentureBeat)
+ AWS says it’s aware of the pitfalls of handing over control to AI. (The Register)
+ The company faces the challenge of building enough infrastructure to support its AI services. (WSJ $)

2 Waymo’s driverless cars are getting surprisingly aggressive
The company’s goal to make the vehicles “confidently assertive” is prompting them to bend the rules. (WSJ $)
+ That said, their cars still have a far lower crash rate than human drivers. (NYT $)

3 The FDA’s top drug regulator has stepped down
After only three weeks in the role. (Ars Technica)+ A leaked vaccine memo from the agency doesn’t inspire confidence. (Bloomberg $)

4 Maybe DOGE isn’t entirely dead after all
Many of its former workers are embedded in various federal agencies. (Wired $)

5 A Chinese startup’s reusable rocket crash-landed after launch
It suffered what it called an “abnormal burn,” scuppering hopes of a soft landing. (Bloomberg $)

6  Startups are building digital clones of major sites to train AI agents
From Amazon to Gmail, they’re creating virtual agent playgrounds. (NYT $)

7 Half of US states now require visitors to porn sites to upload their ID
Missouri has become the 25th state to enact age verification laws. (404 Media)

8 AGI truthers are trying to influence the Pope
They’re desperate for him to take their concerns seriously.(The Verge)
+ How AGI became the most consequential conspiracy theory of our time. (MIT Technology Review)

9 Marketers are leaning into ragebait ads
But does making customers annoyed really translate into sales? (WP $)

10 The surprising role plant pores could play in fighting drought
At night as well as daytime. (Knowable Magazine)
+ Africa fights rising hunger by looking to foods of the past. (MIT Technology Review)

Quote of the day

“Everyone is begging for supply.”

—An anonymous source tells Reuters about the desperate measures Chinese AI companies take to secure scarce chips.

One more thing

The case against humans in space

Elon Musk and Jeff Bezos are bitter rivals in the commercial space race, but they agree on one thing: Settling space is an existential imperative. Space is the place. The final frontier. It is our human destiny to transcend our home world and expand our civilization to extraterrestrial vistas.

This belief has been mainstream for decades, but its rise has been positively meteoric in this new gilded age of astropreneurs.

But as visions of giant orbital stations and Martian cities dance in our heads, a case against human space colonization has found its footing in a number of recent books, from doubts about the practical feasibility of off-Earth communities, to realism about the harsh environment of space and the enormous tax it would exact on the human body. Read the full story.

—Becky Ferreira

We can still have nice things

A place for comfort, fun and distraction to brighten up your day. (Got any ideas? Drop me a line or skeet ’em at me.)

+ This compilation of 21st century floor fillers is guaranteed to make you feel old.
+ A fire-loving amoeba has been found chilling out in volcanic hot springs.
+ This old-school Terminator 2 game is pixel perfection.
+ How truthful an adaptation is your favorite based-on-a-true-story movie? Let’s take a look at the data.

In 1913, Henry Ford cut the time it took to build a Model T from 12 hours to just over 90 minutes. He accomplished this feat through a revolutionary breakthrough in process design: Instead of skilled craftsmen building a car from scratch by hand, Ford created an assembly line where standardized tasks happened in sequence, at scale.

The IT industry is having a similar moment of reinvention. Across operations from software development to cloud migration, organizations are adopting an AI-infused factory model that replaces manual, one-off projects with templated, scalable systems designed for speed and cost-efficiency.

Take VMware migrations as an example. For years, these projects resembled custom production jobs—bespoke efforts that often took many months or even years to complete. Fluctuating licensing costs added a layer of complexity, just as business leaders began pushing for faster modernization to make their organizations AI-ready. That urgency has become nearly universal: According to a recent IDC report, six in 10 organizations evaluating or using cloud services say their IT infrastructure requires major transformation, while 82% report their cloud environments need modernization.

Download the full article.

This content was produced by Insights, the custom content arm of MIT Technology Review. It was not written by MIT Technology Review’s editorial staff.

This content was researched, designed, and written by human writers, editors, analysts, and illustrators. This includes the writing of surveys and collection of data for surveys. AI tools that may have been used were limited to secondary production processes that passed thorough human review.

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology.

The State of AI: Welcome to the economic singularity

—David Rotman and Richard Waters

Any far-reaching new technology is always uneven in its adoption, but few have been more uneven than generative AI. That makes it hard to assess its likely impact on individual businesses, let alone on productivity across the economy as a whole.

At one extreme, AI coding assistants have revolutionized the work of software developers. At the other extreme, most companies are seeing little if any benefit from their initial investments. 

That has provided fuel for the skeptics who maintain that—by its very nature as a probabilistic technology prone to hallucinating—generative AI will never have a deep impact on business. To students of tech history, though, the lack of immediate impact is normal. Read the full story.

If you’re an MIT Technology Review subscriber, you can join David and Richard, alongside our editor in chief, Mat Honan, for an exclusive conversation digging into what’s happening across different markets live on Tuesday, December 9 at 1pm ET.  Register here! 

The State of AI is our subscriber-only collaboration between the Financial Times and MIT Technology Review examining the ways in which AI is reshaping global power. Sign up to receive future editions every Monday.

The must-reads

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

1 DeepSeek has unveiled two new experimental AI models 
DeepSeek-V3.2 is designed to match OpenAI’s GPT-5’s reasoning capabilities. (Bloomberg $)
+ Here’s how DeepSeek slashes its models’ computational burden. (VentureBeat)
+ It’s achieved these results despite its limited access to powerful chips. (SCMP $)

2 OpenAI has issued a “code red” warning to its employees
It’s a call to arms to improve ChatGPT, or risk being overtaken. (The Information $)
+ Both Google and Anthropic are snapping at OpenAI’s heels. (FT $)
+ Advertising and other initiatives will be pushed back to accommodate the new focus. (WSJ $)

3 How to know when the AI bubble has burst
These are the signs to look out for. (Economist $)
+ Things could get a whole lot worse for the economy if and when it pops. (Axios)
+ We don’t really know how the AI investment surge is being financed. (The Guardian)

4 Some US states are making it illegal for AI to discriminate against you
California is the latest to give workers more power to fight algorithms. (WP $)

5 This AI startup is working on a post-transformer future
Transformer architecture underpins the current AI boom—but Pathway is developing something new. (WSJ $)
+ What the next frontier of AI could look like. (IEEE Spectrum)

6 India is demanding smartphone makers install a government app
Which privacy advocates say is unacceptable snooping. (FT $)
+ India’s tech talent is looking for opportunities outside the US. (Rest of World)

7 College students are desperate to sign up for AI majors
AI is now the second-largest major at MIT behind computer science. (NYT $)
+ AI’s giants want to take over the classroom. (MIT Technology Review)

8 America’s musical heritage is at serious risk
Much of it is stored on studio tapes, which are deteriorating over time. (NYT $)
+ The race to save our online lives from a digital dark age. (MIT Technology Review)

9 Celebrities are increasingly turning on AI
That doesn’t stop fans from casting them in slop videos anyway. (The Verge)

10 Samsung has revealed its first tri-folding phone
But will people actually want to buy it? (Bloomberg $)
+ It’ll cost more than $2,000 when it goes on sale in South Korea. (Reuters)

Quote of the day

“The Chinese will not pause. They will take over.”

—Michael Lohscheller, chief executive of Swedish electric car maker Polestar, tells the Guardian why Europe should stick to its plan to ban the production of new petrol and diesel cars by 2035. 

One more thing

Inside Amsterdam’s high-stakes experiment to create fair welfare AI

Amsterdam thought it was on the right track. City officials in the welfare department believed they could build technology that would prevent fraud while protecting citizens’ rights. They followed these emerging best practices and invested a vast amount of time and money in a project that eventually processed live welfare applications. But in their pilot, they found that the system they’d developed was still not fair and effective. Why?

Lighthouse Reports, MIT Technology Review, and the Dutch newspaper Trouw have gained unprecedented access to the system to try to find out. Read about what we discovered.

—Eileen Guo, Gabriel Geiger & Justin-Casimir Braun

We can still have nice things

A place for comfort, fun and distraction to brighten up your day. (Got any ideas? Drop me a line or skeet ’em at me.)

+ Hear me out: a truly great festive film doesn’t need to be about Christmas at all.
+ Maybe we should judge a book by its cover after all.
+ Happy birthday to Ms Britney Spears, still the princess of pop at 44!
+ The fascinating psychology behind why we love travelling so much.

Welcome back to The State of AI, a new collaboration between the Financial Times and MIT Technology Review. Every Monday for the next two weeks, writers from both publications will debate one aspect of the generative AI revolution reshaping global power.

This week, Richard Waters, FT columnist and former West Coast editor, talks with MIT Technology Review’s editor at large David Rotman about the true impact of AI on the job market.

Bonus: If you’re an MIT Technology Review subscriber, you can join David and Richard, alongside MIT Technology Review’s editor in chief, Mat Honan, for an exclusive conversation live on Tuesday, December 9 at 1pm ET about this topic. Sign up to be a part here.

Richard Waters writes:

Any far-reaching new technology is always uneven in its adoption, but few have been more uneven than generative AI. That makes it hard to assess its likely impact on individual businesses, let alone on productivity across the economy as a whole.

At one extreme, AI coding assistants have revolutionized the work of software developers. Mark Zuckerberg recently predicted that half of Meta’s code would be written by AI within a year. At the other extreme, most companies are seeing little if any benefit from their initial investments. A widely cited study from MIT found that so far, 95% of gen AI projects produce zero return.

That has provided fuel for the skeptics who maintain that—by its very nature as a probabilistic technology prone to hallucinating—generative AI will never have a deep impact on business.

To many students of tech history, though, the lack of immediate impact is just the normal lag associated with transformative new technologies. Erik Brynjolfsson, then an assistant professor at MIT, first described what he called the “productivity paradox of IT” in the early 1990s. Despite plenty of anecdotal evidence that technology was changing the way people worked, it wasn’t showing up in the aggregate data in the form of higher productivity growth. Brynjolfsson’s conclusion was that it just took time for businesses to adapt.

Big investments in IT finally showed through with a notable rebound in US productivity growth starting in the mid-1990s. But that tailed off a decade later and was followed by a second lull.

In the case of AI, companies need to build new infrastructure (particularly data platforms), redesign core business processes, and retrain workers before they can expect to see results. If a lag effect explains the slow results, there may at least be reasons for optimism: Much of the cloud computing infrastructure needed to bring generative AI to a wider business audience is already in place.

The opportunities and the challenges are both enormous. An executive at one Fortune 500 company says his organization has carried out a comprehensive review of its use of analytics and concluded that its workers, overall, add little or no value. Rooting out the old software and replacing that inefficient human labor with AI might yield significant results. But, as this person says, such an overhaul would require big changes to existing processes and take years to carry out.

There are some early encouraging signs. US productivity growth, stuck at 1% to 1.5% for more than a decade and a half, rebounded to more than 2% last year. It probably hit the same level in the first nine months of this year, though the lack of official data due to the recent US government shutdown makes this impossible to confirm.

It is impossible to tell, though, how durable this rebound will be or how much can be attributed to AI. The effects of new technologies are seldom felt in isolation. Instead, the benefits compound. AI is riding earlier investments in cloud and mobile computing. In the same way, the latest AI boom may only be the precursor to breakthroughs in fields that have a wider impact on the economy, such as robotics. ChatGPT might have caught the popular imagination, but OpenAI’s chatbot is unlikely to have the final word.

David Rotman replies: 

This is my favorite discussion these days when it comes to artificial intelligence. How will AI affect overall economic productivity? Forget about the mesmerizing videos, the promise of companionship, and the prospect of agents to do tedious everyday tasks—the bottom line will be whether AI can grow the economy, and that means increasing productivity. 

But, as you say, it’s hard to pin down just how AI is affecting such growth or how it will do so in the future. Erik Brynjolfsson predicts that, like other so-called general purpose technologies, AI will follow a J curve in which initially there is a slow, even negative, effect on productivity as companies invest heavily in the technology before finally reaping the rewards. And then the boom. 

But there is a counterexample undermining the just-be-patient argument. Productivity growth from IT picked up in the mid-1990s but since the mid-2000s has been relatively dismal. Despite smartphones and social media and apps like Slack and Uber, digital technologies have done little to produce robust economic growth. A strong productivity boost never came.

Daron Acemoglu, an economist at MIT and a 2024 Nobel Prize winner, argues that the productivity gains from generative AI will be far smaller and take far longer than AI optimists think. The reason is that though the technology is impressive in many ways, the field is too narrowly focused on products that have little relevance to the largest business sectors.

The statistic you cite that 95% of AI projects lack business benefits is telling. 

Take manufacturing. No question, some version of AI could help; imagine a worker on the factory floor snapping a picture of a problem and asking an AI agent for advice. The problem is that the big tech companies creating AI aren’t really interested in solving such mundane tasks, and their large foundation models, mostly trained on the internet, aren’t all that helpful. 

It’s easy to blame the lack of productivity impact from AI so far on business practices and poorly trained workers. Your example of the executive of the Fortune 500 company sounds all too familiar. But it’s more useful to ask how AI can be trained and fine-tuned to give workers, like nurses and teachers and those on the factory floor, more capabilities and make them more productive at their jobs. 

The distinction matters. Some companies announcing large layoffs recently cited AI as the reason. The worry, however, is that it’s just a short-term cost-saving scheme. As economists like Brynjolfsson and Acemoglu agree, the productivity boost from AI will come when it’s used to create new types of jobs and augment the abilities of workers, not when it is used just to slash jobs to reduce costs. 

Richard Waters responds : 

I see we’re both feeling pretty cautious, David, so I’ll try to end on a positive note. 

Some analyses assume that a much greater share of existing work is within the reach of today’s AI. McKinsey reckons 60% (versus 20% for Acemoglu) and puts annual productivity gains across the economy at as much as 3.4%. Also, calculations like these are based on automation of existing tasks; any new uses of AI that enhance existing jobs would, as you suggest, be a bonus (and not just in economic terms).

Cost-cutting always seems to be the first order of business with any new technology. But we’re still in the early stages and AI is moving fast, so we can always hope.

Further reading

FT chief economics commentator Martin Wolf has been skeptical about whether tech investment boosts productivity but says AI might prove him wrong. The downside: Job losses and wealth concentration might lead to “techno-feudalism.”

The FT‘s Robert Armstrong argues that the boom in data center investment need not turn to bust. The biggest risk is that debt financing will come to play too big a role in the buildout.

Last year, David Rotman wrote for MIT Technology Review about how we can make sure AI works for us in boosting productivity, and what course corrections will be required.

David also wrote this piece about how we can best measure the impact of basic R&D funding on economic growth, and why it can often be bigger than you might think.

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology.

An AI model trained on prison phone calls now looks for planned crimes in those calls

A US telecom company trained an AI model on years of inmates’ phone and video calls and is now piloting that model to scan their calls, texts, and emails in the hope of predicting and preventing crimes.

Securus Technologies president Kevin Elder told MIT Technology Review that the company began building its AI tools in 2023, using its massive database of recorded calls to train AI models to detect criminal activity. It created one model, for example, using seven years of calls made by inmates in the Texas prison system, but it has been working on models for other states and counties.

However, prisoner rights advocates say that the new AI system enables a system of invasive surveillance, and courts have specified few limits to this power.  Read the full story.

—James O’Donnell

Nominations are now open for our global 2026 Innovators Under 35 competition

We have some exciting news: Nominations are now open for MIT Technology Review’s 2026 Innovators Under 35 competition. This annual list recognizes 35 of the world’s best young scientists and inventors, and our newsroom has produced it for more than two decades. 

It’s free to nominate yourself or someone you know, and it only takes a few moments. Here’s how to submit your nomination.

The must-reads

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

1 New York is cracking down on personalized pricing algorithms
A new law forces retailers to declare if their pricing is informed by users’ data. (NYT $)
+ The US National Retail Federation tried to block it from passing. (TechCrunch)

2 The White House has launched a media bias tracker
Complete with a “media offender of the week” section and a Hall of Shame. (WP $)
+ The Washington Post is currently listed as the site’s top offender. (The Guardian)
+ Donald Trump has lashed out at several reporters in the past few weeks. (The Hill)

3 American startups are hooked on open-source Chinese AI models
They’re cheap and customizable—what’s not to like? (NBC News)
+ Americans also love China’s cheap goods, regardless of tariffs. (WP $)
+ The State of AI: Is China about to win the race? (MIT Technology Review)

4 How police body cam footage became viral YouTube content
Recent arrestees live in fear of ending up on popular channels. (Vox)
+ AI was supposed to make police bodycams better. What happened? (MIT Technology Review)

5 Construction workers are cashing in on the data center boom
Might as well enjoy it while it lasts. (WSJ $)
+ The data center boom in the desert. (MIT Technology Review)

6 China isn’t convinced by crypto
Even though bitcoin mining is quietly making a (banned) comeback. (Reuters)
+ The country’s central bank is no fan of stablecoins. (CoinDesk)

7 A startup is treating its AI companions like characters in a novel
Could that approach make for better AI companions? (Fast Company $)
+ Gemini is the most empathetic model, apparently. (Semafor)
+ The looming crackdown on AI companionship. (MIT Technology Review)

8 Ozempic is so yesterday
New weight-loss drugs are tailored to individual patients. (The Atlantic $)
+ What we still don’t know about weight-loss drugs. (MIT Technology Review)

9 AI is upending how consultants work
For the third year in a row, big firms are freezing junior workers’ salaries. (FT $)

10 Behind the scenes of Disney’s AI animation accelerator
What took five months to create has been whittled down to under five weeks. (CNET)
+ Director supremo James Cameron appears to have changed his mind about AI. (TechCrunch)
+ Why are people scrolling through weirdly-formatted TV clips? (WP $)

Quote of the day

“[I hope AI] comes to a point where it becomes sort of mental junk food and we feel sick and we don’t know why.”

—Actor Jenna Ortega outlines her hopes for AI’s future role in filmmaking, Variety reports.

One more thing

The weeds are winning

Since the 1980s, more and more plants have evolved to become immune to the biochemical mechanisms that herbicides leverage to kill them. This herbicidal resistance threatens to decrease yields—out-of-control weeds can reduce them by 50% or more, and extreme cases can wipe out whole fields.

At worst, it can even drive farmers out of business. It’s the agricultural equivalent of antibiotic resistance, and it keeps getting worse. Weeds have evolved resistance to 168 different herbicides and 21 of the 31 known “modes of action,” which means the specific biochemical target or pathway a chemical is designed to disrupt.

Agriculture needs to embrace a diversity of weed control practices. But that’s much easier said than done. Read the full story.

—Douglas Main

We can still have nice things

A place for comfort, fun and distraction to brighten up your day. (Got any ideas? Drop me a line or skeet ’em at me.)

+ Now we’re finally in December, don’t let Iceland’s gigantic child-eating Yule Cat give you nightmares
+ These breathtaking sculpture parks are serious must-sees ($)
+ 1985 sure was a vintage year for films.
+ Is nothing sacred?! Now Ozempic has come for our Christmas trees!

We have some exciting news: Nominations are now open for MIT Technology Review’s 2026 Innovators Under 35 competition. This annual list recognizes 35 of the world’s best young scientists and inventors, and our newsroom has produced it for more than two decades. 

It’s free to nominate yourself or someone you know, and it only takes a few moments. Submit your nomination before 5 p.m. ET on Tuesday, January 20, 2026. 

We’re looking for people who are making important scientific discoveries and applying that knowledge to build new technologies. Or those who are engineering new systems and algorithms that will aid our work or extend our abilities. 

Each year, many honorees are focused on improving human health or solving major problems like climate change; others are charting the future path of artificial intelligence or developing the next generation of robots. 

The most successful candidates will have made a clear advance that is expected to have a positive impact beyond their own field. They should be the primary scientific or technical driver behind the work involved, and we like to see some signs that a candidate’s innovation is gaining real traction. You can look at last year’s list to get an idea of what we look out for.

We encourage self-nominations, and if you previously nominated someone who wasn’t selected, feel free to put them forward again. Please note: To be eligible for the 2026 list, nominees must be under the age of 35 as of October 1, 2026. 

Semifinalists will be notified by early March and asked to complete an application at that time. Winners are then chosen by the editorial staff of MIT Technology Review, with input from a panel of expert judges. (Here’s more info about our selection process and timelines.) 

If you have any questions, please contact tr35@technologyreview.com. We look forward to reviewing your nominations. Good luck! 

A US telecom company trained an AI model on years of inmates’ phone and video calls and is now piloting that model to scan their calls, texts, and emails in the hope of predicting and preventing crimes. 

Securus Technologies president Kevin Elder told MIT Technology Review that the company began building its AI tools in 2023, using its massive database of recorded calls to train AI models to detect criminal activity. It created one model, for example, using seven years of calls made by inmates in the Texas prison system, but it has been working on building other state- or county-specific models.

Over the past year, Elder says, Securus has been piloting the AI tools to monitor inmate conversations in real time. The company declined to specify where this is taking place, but its customers include jails holding people awaiting trial and prisons for those serving sentences. Some of these facilities using Securus technology also have agreements with Immigrations and Customs Enforcement to detain immigrants, though Securus does not contract with ICE directly.

“We can point that large language model at an entire treasure trove [of data],” Elder says, “to detect and understand when crimes are being thought about or contemplated, so that you’re catching it much earlier in the cycle.”

As with its other monitoring tools, investigators at detention facilities can deploy the AI features to monitor randomly selected conversations or those of individuals suspected by facility investigators of criminal activity, according to Elder. The model will analyze phone and video calls, text messages, and emails and then flag sections for human agents to review. These agents then send them to investigators for follow-up. 

In an interview, Elder said Securus’ monitoring efforts have helped disrupt human trafficking and gang activities organized from within prisons, among other crimes, and said its tools are also used to identify prison staff who are bringing in contraband. But the company did not provide MIT Technology Review with any cases specifically uncovered by its new AI models. 

People in prison, and those they call, are notified that their conversations are recorded. But this doesn’t mean they’re aware that those conversations could be used to train an AI model, says Bianca Tylek, executive director of the prison rights advocacy group Worth Rises. 

“That’s coercive consent; there’s literally no other way you can communicate with your family,” Tylek says. And since inmates in the vast majority of states pay for these calls, she adds, “not only are you not compensating them for the use of their data, but you’re actually charging them while collecting their data.”

A Securus spokesperson said the use of data to train the tool “is not focused on surveilling or targeting specific individuals, but rather on identifying broader patterns, anomalies, and unlawful behaviors across the entire communication system.” They added that correctional facilities determine their own recording and monitoring policies, which Securus follows, and did not directly answer whether inmates can opt out of having their recordings used to train AI.

Other advocates for inmates say Securus has a history of violating their civil liberties. For example, leaks of its recordings databases showed the company had improperly recorded thousands of calls between inmates and their attorneys. Corene Kendrick, the deputy director of the ACLU’s National Prison Project, says that the new AI system enables a system of invasive surveillance, and courts have specified few limits to this power.

“[Are we] going to stop crime before it happens because we’re monitoring every utterance and thought of incarcerated people?” Kendrick says. “I think this is one of many situations where the technology is way far ahead of the law.”

The company spokesperson said the tool’s function is to make monitoring more efficient amid staffing shortages, “not to surveil individuals without cause.”

Securus will have an easier time funding its AI tool thanks to the company’s recent win in a battle with regulators over how telecom companies can spend the money they collect from inmates’ calls.

In 2024, the Federal Communications Commission issued a major reform, shaped and lauded by advocates for prisoners’ rights, that forbade telecoms from passing the costs of recording and surveilling calls on to inmates. Companies were allowed to continue to charge inmates a capped rate for calls, but prisons and jails were ordered to pay for most security costs out of their own budgets.

Negative reactions to this change were swift. Associations of sheriffs (who typically run county jails) complained they could no longer afford proper monitoring of calls, and attorneys general from 14 states sued over the ruling. Some prisons and jails warned they would cut off access to phone calls. 

While it was building and piloting its AI tool, Securus held meetings with the FCC and lobbied for a rule change, arguing that the 2024 reform went too far and asking that the agency again allow companies to use fees collected from inmates to pay for security. 

In June, Brendan Carr, whom President Donald Trump appointed to lead the FCC, said it would postpone all deadlines for jails and prisons to adopt the 2024 reforms, and even signaled that the agency wants to help telecom companies fund their AI surveillance efforts with the fees paid by inmates. In a press release, Carr wrote that rolling back the 2024 reforms would “lead to broader adoption of beneficial public safety tools that include advanced AI and machine learning.”

On October 28, the agency went further: It voted to pass new, higher rate caps and allow companies like Securus to pass security costs relating to recording and monitoring of calls—like storing recordings, transcribing them, or building AI tools to analyze such calls, for example—on to inmates. A spokesperson for Securus told MIT Technology Review that the company aims to balance affordability with the need to fund essential safety and security tools. “These tools, which include our advanced monitoring and AI capabilities, are fundamental to maintaining secure facilities for incarcerated individuals and correctional staff and to protecting the public,” they wrote.

FCC commissioner Anna Gomez dissented in last month’s ruling. “Law enforcement,” she wrote in a statement, “should foot the bill for unrelated security and safety costs, not the families of incarcerated people.”

The FCC will be seeking comment on these new rules before they take final effect. 

This story was updated on December 2 to clarify that Securus does not contract with ICE facilities.

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology.

What we still don’t know about weight-loss drugs

Weight-loss drugs have been back in the news this week. First, we heard that Eli Lilly, the company behind Mounjaro and Zepbound, became the first healthcare company in the world to achieve a trillion-dollar valuation.

But we also learned that, disappointingly, GLP-1 drugs don’t seem to help people with Alzheimer’s disease. And that people who stop taking the drugs when they become pregnant can experience potentially dangerous levels of weight gain. On top of that, some researchers worry that people are using the drugs postpartum to lose pregnancy weight without understanding potential risks.

All of this news should serve as a reminder that there’s a lot we still don’t know about these drugs. So let’s look at the enduring questions surrounding GLP-1 agonist drugs.

—Jessica Hamzelou

This article first appeared in The Checkup, MIT Technology Review’s weekly biotech newsletter. To receive it in your inbox every Thursday, and read articles like this first, sign up here.

If you’re interested in weight loss drugs and how they affect us, take a look at:

+ GLP-1 agonists like Wegovy, Ozempic, and Mounjaro might benefit heart and brain health—but research suggests they might also cause pregnancy complications and harm some users. Read the full story.

+ We’ve never understood how hunger works. That might be about to change. Read the full story.

+ Weight-loss injections have taken over the internet. But what does this mean for people IRL?

+ This vibrating weight-loss pill seems to work—in pigs. Read the full story.

What we know about how AI is affecting the economy

There’s a lot at stake when it comes to understanding how AI is changing the economy right now. Should we be pessimistic? Optimistic? Or is the situation too nuanced for that?

Hopefully, we can point you towards some answers. Mat Honan, our editor in chief, will hold a special subscriber-only Roundtables conversation with our editor at large David Rotman, and Richard Waters, Financial Times columnist, exploring what’s happening across different markets. Register here to join us at 1pm ET on Tuesday December 9.

The event is part of the Financial Times and MIT Technology Review “The State of AI” partnership, exploring the global impact of artificial intelligence. Over the past month, we’ve been running discussions between our journalists—sign up here to receive future editions every Monday.

The must-reads

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

1 Tech billionaires are gearing up to fight AI regulation 
By amassing multi-million dollar war chests ahead of the 2026 US midterm elections. (WSJ $)
+ Donald Trump’s “Manhattan Project” for AI is certainly ambitious. (The Information $)

2 The EU wants to hold social media platforms liable for financial scams
New rules will force tech firms to compensate banks if they fail to remove reported scams. (Politico)

3 China is worried about a humanoid robot bubble
Because more than 150 companies there are building very similar machines. (Bloomberg $)
+ It could learn some lessons from the current AI bubble. (CNN)+ Why the humanoid workforce is running late. (MIT Technology Review)

4 A Myanmar scam compound was blown up
But its residents will simply find new bases for their operations. (NYT $)
+ Experts suspect the destruction may have been for show. (Wired $)
+ Inside a romance scam compound—and how people get tricked into being there. (MIT Technology Review)

5 Navies across the world are investing in submarine drones 
They cost a fraction of what it takes to run a traditional manned sub. (The Guardian)
+ How underwater drones could shape a potential Taiwan-China conflict. (MIT Technology Review)

6 What to expect from China’s seemingly unstoppable innovation drive
Its extremely permissive regulators play a big role. (Economist $)
+ Is China about to win the AI race? (MIT Technology Review)

7 The UK is waging a war on VPNs
Good luck trying to persuade people to stop using them. (The Verge)

8 We’re learning more about Jeff Bezos’ mysterious clock project
He’s backed the Clock of the Long Now for years—and construction is amping up. (FT $)
+ How aging clocks can help us understand why we age—and if we can reverse it. (MIT Technology Review)

9 Have we finally seen the first hints of dark matter?
These researchers seem to think so. (New Scientist $)

10 A helpful robot is helping archaeologists reconstruct Pompeii
Reassembling ancient frescos is fiddly and time-consuming, but less so if you’re a dextrous machine. (Reuters)

Quote of the day

“We do fail… a lot.”

—Defense company Anduril explains its move-fast-and-break-things ethos to the Wall Street Journal in response to reports its systems have been marred by issues in Ukraine.

One more thing

How to build a better AI benchmark

It’s not easy being one of Silicon Valley’s favorite benchmarks.

SWE-Bench (pronounced “swee bench”) launched in November 2024 as a way to evaluate an AI model’s coding skill. It has since quickly become one of the most popular tests in AI. A SWE-Bench score has become a mainstay of major model releases from OpenAI, Anthropic, and Google—and outside of foundation models, the fine-tuners at AI firms are in constant competition to see who can rise above the pack.

Despite all the fervor, this isn’t exactly a truthful assessment of which model is “better.” Entrants have begun to game the system—which is pushing many others to wonder whether there’s a better way to actually measure AI achievement. Read the full story.

—Russell Brandom

We can still have nice things

A place for comfort, fun and distraction to brighten up your day. (Got any ideas? Drop me a line or skeet ’em at me.)

+ Aww, these sharks appear to be playing with pool toys.
+ Strange things are happening over on Easter Island (even weirder than you can imagine)
+ Very cool—archaeologists have uncovered a Roman tomb that’s been sealed shut for 1,700 years.
+ This Japanese mass media collage is making my eyes swim, in a good way.

MIT Technology Review Explains: Let our writers untangle the complex, messy world of technology to help you understand what’s coming next. You can read more from the series here.

Weight-loss drugs have been back in the news this week. First, we heard that Eli Lilly, the company behind the drugs Mounjaro and Zepbound, became the first healthcare company in the world to achieve a trillion-dollar valuation.

Those two drugs, which are prescribed for diabetes and obesity respectively, are generating billions of dollars in revenue for the company. Other GLP-1 agonist drugs—a class that includes Mounjaro and Zepbound, which have the same active ingredient—have also been approved to reduce the risk of heart attack and stroke in overweight people. Many hope these apparent wonder drugs will also treat neurological disorders and potentially substance use disorders, too.

But this week we also learned that, disappointingly, GLP-1 drugs don’t seem to help people with Alzheimer’s disease. And that people who stop taking the drugs when they become pregnant can experience potentially dangerous levels of weight gain during their pregnancies. On top of that, some researchers worry that people are using the drugs postpartum to lose pregnancy weight without understanding potential risks.

All of this news should serve as a reminder that there’s a lot we still don’t know about these drugs. This week, let’s look at the enduring questions surrounding GLP-1 agonist drugs.

First a quick recap. Glucagon-like peptide-1 is a hormone made in the gut that helps regulate blood sugar levels. But we’ve learned that it also appears to have effects across the body. Receptors that GLP-1 can bind to have been found in multiple organs and throughout the brain, says Daniel Drucker, an endocrinologist at the University of Toronto who has been studying the hormone for decades.

GLP-1 agonist drugs essentially mimic the hormone’s action. Quite a few have been developed, including semaglutide, tirzepatide, liraglutide, and exenatide, which have brand names like Ozempic, Saxenda and Wegovy. Some of them are recommended for some people with diabetes.

But because these drugs also seem to suppress appetite, they have become hugely popular weight loss aids. And studies have found that many people who take them for diabetes or weight loss experience surprising side effects; that their mental health improves, for example, or that they feel less inclined to smoke or consume alcohol. Research has also found that the drugs seem to increase the growth of brain cells in lab animals.

So far, so promising. But there are a few outstanding gray areas.

Are they good for our brains?

Novo Nordisk, a competitor of Eli Lilly, manufactures GLP-1 drugs Wegovy and Saxenda. The company recently trialed an oral semaglutide in people with Alzheimer’s disease who had mild cognitive impairment or mild dementia. The placebo-controlled trial included 3808 volunteers.

Unfortunately, the company found that the drug did not appear to delay the progression of Alzheimer’s disease in the volunteers who took it.

The news came as a huge disappointment to the research community. “It was kind of crushing,” says Drucker. That’s despite the fact that, deep down, he wasn’t expecting a “clear win.” Alzheimer’s disease has proven notoriously difficult to treat, and by the time people get a diagnosis, a lot of damage has already taken place.

But he is one of many that isn’t giving up hope entirely. After all, research suggests that GLP-1 reduces inflammation in the brain and improves the health of neurons, and that it appears to improve the way brain regions communicate with each other. This all implies that GLP-1 drugs should benefit the brain, says Drucker. There’s still a chance that the drugs might help stave off Alzheimer’s in those who are still cognitively healthy.

Are they safe before, during or after pregnancy?

Other research published this week raises questions about the effects of GLP-1s taken around the time of pregnancy. At the moment, people are advised to plan to stop taking the medicines two months before they become pregnant. That’s partly because some animal studies suggest the drugs can harm the development of a fetus, but mainly because scientists haven’t studied the impact on pregnancy in humans.

Among the broader population, research suggests that many people who take GLP-1s for weight loss regain much of their lost weight once they stop taking those drugs. So perhaps it’s not surprising that a study published in JAMA earlier this week saw a similar effect in pregnant people.

The study found that people who had been taking those drugs gained around 3.3kg more than others who had not. And those who had been taking the drugs also appeared to have a slightly higher risk of gestational diabetes, blood pressure disorders and even preterm birth.

It sounds pretty worrying. But a different study published in August had the opposite finding—it noted a reduction in the risk of those outcomes among women who had taken the drugs before becoming pregnant.

If you’re wondering how to make sense of all this, you’re not the only one. No one really knows how these drugs should be used before pregnancy—or during it for that matter.

Another study out this week found that people (in Denmark) are increasingly taking GLP-1s postpartum to lose weight gained during pregnancy. Drucker tells me that, anecdotally, he gets asked about this potential use a lot.

But there’s a lot going on in a postpartum body. It’s a time of huge physical and hormonal change that can include bonding, breastfeeding and even a rewiring of the brain. We have no idea if, or how, GLP-1s might affect any of those.

How—and when—can people safely stop using them?

Yet another study out this week—you can tell GLP-1s are one of the hottest topics in medicine right now—looked at what happens when people stop taking tirzepatide (marketed as Zepbound) for their obesity.

The trial participants all took the drug for 36 weeks, at which point half continued with the drug, and half were switched to a placebo for another 52 weeks. During that first 36 weeks, the weight and heart health of the participants improved.

But by the end of the study, most of those that had switched to a placebo had regained more than 25% of the weight they had originally lost. One in four had regained more than 75% of that weight, and 9% ended up at a higher weight than when they’d started the study. Their heart health also worsened.

Does that mean that people need to take these drugs forever? Scientists don’t have the answer to that one, either. Or if taking the drugs indefinitely is safe. The answer might depend on the individual, their age or health status, or what they are using the drug for.

There are other gray areas. GLP-1s look promising for substance use disorders, but we don’t yet know how effective they might be. We don’t know the long-term effects these drugs have on children who take them. And we don’t know the long-term consequences these drugs might have for healthy-weight people who take them for weight loss.

Earlier this year, Drucker accepted a Breakthrough Prize in Life Sciences at a glitzy event in California. “All of these Hollywood celebrities were coming up to me and saying ‘thank you so much,’” he says. “A lot of these people don’t need to be on these medicines.”

This article first appeared in The Checkup, MIT Technology Review’s weekly biotech newsletter. To receive it in your inbox every Thursday, and read articles like this first, sign up here.

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology.

This year’s UN climate talks avoided fossil fuels, again

Over the past few weeks in Belem, Brazil, attendees of this year’s UN climate talks dealt with oppressive heat and flooding, and at one point a literal fire broke out, delaying negotiations. The symbolism was almost too much to bear.

While many, including the president of Brazil, framed this year’s conference as one of action, the talks ended with a watered-down agreement. The final draft doesn’t even include the phrase “fossil fuels.”

As emissions and global temperatures reach record highs again this year, I’m left wondering: Why is it so hard to formally acknowledge what’s causing the problem?

—Casey Crownhart

This article is from The Spark, MIT Technology Review’s weekly climate newsletter. To receive it in your inbox every Wednesday, sign up here.

New noninvasive endometriosis tests are on the rise

Endometriosis inflicts debilitating pain and heavy bleeding on more than 11% of reproductive-­age women in the United States. Diagnosis takes nearly 10 years on average, partly because half the cases don’t show up on scans, and surgery is required to obtain tissue samples.

But a new generation of noninvasive tests are emerging that could help accelerate diagnosis and improve management of this poorly understood condition. Read the full story.

—Colleen de Bellefonds

This story is from the last print issue of MIT Technology Review magazine, which is full of fascinating stories about the body. If you haven’t already, subscribe now to receive future issues once they land.

The must-reads

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

1 OpenAI claims a teenager circumvented its safety features before ending his life
It says ChatGPT directed Adam Raine to seek help more than 100 times. (TechCrunch)
+ OpenAI is strongly refuting the idea it’s liable for the 16-year old’s death. (NBC News)
+ The looming crackdown on AI companionship. (MIT Technology Review)

2 The CDC’s new deputy director prefers natural immunity to vaccines
And he wasn’t even the worst choice among those considered for the role. (Ars Technica)
+ Meet Jim O’Neill, the longevity enthusiast who is now RFK Jr.’s right-hand man. (MIT Technology Review)

3 An MIT study says AI could already replace 12% of the US workforce
Researchers drew that conclusion after simulating a digital twin of the US labor market. (CNBC)
+ Separate research suggests it could replace 3 million jobs in the UK, too. (The Guardian)
+ AI usage looks unlikely to keep climbing. (Economist $)

4 An Italian defense group has created an AI-powered air shield system
It claims the system allows defenders to generate dome-style missile shields. (FT $)
+ Why Trump’s “golden dome” missile defense idea is another ripped straight from the movies. (MIT Technology Review)

5 The EU is considering a ban on social media for under-16s
Following in Australia’s footsteps, whose own ban comes into power next month. (Politico)
+ The European Parliament wants parents to decide on access. (The Guardian)

6 Why do so many astronauts keep getting stuck in space?
America, Russia and now China have had to contend with this situation. (WP $)
+ A rescue craft for three stranded Chinese astronauts has successfully reached them. (The Register)

7 Uploading pictures of your hotel room could help trafficking victims
A new app uses computer vision to determine where pictures of generic-looking rooms were taken. (IEEE Spectrum)

8 This browser tool turns back the clock to a pre-AI slop web
Back to the golden age of pre-November 30 2022. (404 Media)
+ The White House’s slop posts are shockingly bad. (NY Mag $)
+ Animated neo-Nazi propaganda is freely available on X. (The Atlantic $)

9 Grok’s “epic roasts” are as tragic as you’d expect
Test it out at parties at your own peril. (Wired $)

10 Startup founders dread explaining their jobs at Thanksgiving
Yes Grandma, I work with computers. (Insider $)

Quote of the day

“AI cannot ever replace the unique gift that you are to the world.”

—Pope Leo XIV warns students about the dangers of over-relying on AI, New York Magazine reports.

One more thing

Why we should thank pigeons for our AI breakthroughs

People looking for precursors to artificial intelligence often point to science fiction or thought experiments like the Turing test. But an equally important, if surprising and less appreciated, forerunner is American psychologist B.F. Skinner’s research with pigeons in the middle of the 20th century.

Skinner believed that association—learning, through trial and error, to link an action with a punishment or reward—was the building block of every behavior, not just in pigeons but in all living organisms, including human beings.

His “behaviorist” theories fell out of favor in the 1960s but were taken up by computer scientists who eventually provided the foundation for many of the leading AI tools. Read the full story.

—Ben Crair

We can still have nice things

A place for comfort, fun and distraction to brighten up your day. (Got any ideas? Drop me a line or skeet ’em at me.)

+ I hope you had a happy, err, Green Wednesday if you partook this year.
+ Here how to help an endangered species from the comfort of your own home.
+ Polly wants to FaceTime—now! (thanks Alice!)
+ I need Macaulay Culkin’s idea for another Home Alone sequel to get greenlit, stat.

If we didn’t have pictures and videos, I almost wouldn’t believe the imagery that came out of this year’s UN climate talks.

Over the past few weeks in Belem, Brazil, attendees dealt with oppressive heat and flooding, and at one point a literal fire broke out, delaying negotiations. The symbolism was almost too much to bear.

While many, including the president of Brazil, framed this year’s conference as one of action, the talks ended with a watered-down agreement. The final draft doesn’t even include the phrase “fossil fuels.”

As emissions and global temperatures reach record highs again this year, I’m left wondering: Why is it so hard to formally acknowledge what’s causing the problem?

This is the 30th time that leaders have gathered for the Conference of the Parties, or COP, an annual UN conference focused on climate change. COP30 also marks 10 years since the gathering that produced the Paris Agreement, in which world powers committed to limiting global warming to “well below” 2.0 °C above preindustrial levels, with a goal of staying below the 1.5 °C mark. (That’s 3.6 °F and 2.7 °F, respectively, for my fellow Americans.)

Before the conference kicked off this year, host country Brazil’s president, Luiz Inácio Lula da Silva, cast this as the “implementation COP” and called for negotiators to focus on action, and specifically to deliver a road map for a global transition away from fossil fuels.

The science is clear—burning fossil fuels emits greenhouse gases and drives climate change. Reports have shown that meeting the goal of limiting warming to 1.5 °C would require stopping new fossil-fuel exploration and development.

The problem is, “fossil fuels” might as well be a curse word at global climate negotiations. Two years ago, fights over how to address fossil fuels brought talks at COP28 to a standstill. (It’s worth noting that the conference was hosted in Dubai in the UAE, and the leader was literally the head of the country’s national oil company.)

The agreement in Dubai ended up including a line that called on countries to transition away from fossil fuels in energy systems. It was short of what many advocates wanted, which was a more explicit call to phase out fossil fuels entirely. But it was still hailed as a win. As I wrote at the time: “The bar is truly on the floor.”

And yet this year, it seems we’ve dug into the basement.

At one point about 80 countries, a little under half of those present, demanded a concrete plan to move away from fossil fuels.

But oil producers like Saudi Arabia were insistent that fossil fuels not be singled out. Other countries, including some in Africa and Asia, also made a very fair point: Western nations like the US have burned the most fossil fuels and benefited from it economically. This contingent maintains that legacy polluters have a unique responsibility to finance the transition for less wealthy and developing nations rather than simply barring them from taking the same development route. 

The US, by the way, didn’t send a formal delegation to the talks, for the first time in 30 years. But the absence spoke volumes. In a statement to the New York Times that sidestepped the COP talks, White House spokesperson Taylor Rogers said that president Trump had “set a strong example for the rest of the world” by pursuing new fossil-fuel development.

To sum up: Some countries are economically dependent on fossil fuels, some don’t want to stop depending on fossil fuels without incentives from other countries, and the current US administration would rather keep using fossil fuels than switch to other energy sources. 

All those factors combined help explain why, in its final form, COP30’s agreement doesn’t name fossil fuels at all. Instead, there’s a vague line that leaders should take into account the decisions made in Dubai, and an acknowledgement that the “global transition towards low greenhouse-gas emissions and climate-resilient development is irreversible and the trend of the future.”

Hopefully, that’s true. But it’s concerning that even on the world’s biggest stage, naming what we’re supposed to be transitioning away from and putting together any sort of plan to actually do it seems to be impossible.

This article is from The Spark, MIT Technology Review’s weekly climate newsletter. To receive it in your inbox every Wednesday, sign up here.

Today’s IT leaders face competing mandates to do more (“make us an ‘AI-first’ enterprise—yesterday”) with less (“no new hires for at least the next six months”).

VMware has become a focal point of these dueling directives. It remains central to enterprise IT, with 80% of organizations using VMware infrastructure products. But shifting licensing models are prompting teams to reconsider how they manage and scale these workloads, often on tighter budgets.

For many organizations, the path forward involves adopting a LessOps model, an operational strategy that makes hybrid environments manageable without increasing headcount. This operational philosophy minimizes human intervention through extensive automation and selfservice capabilities while maintaining governance and compliance.

In practice, VMware-to-cloud migrations create a “two birds, one stone” opportunity. They present a practical moment to codify the automation and governance practices LessOps depends on—laying the groundwork for a leaner, more resilient IT operating model.

Download the full article.

This content was produced by Insights, the custom content arm of MIT Technology Review. It was not written by MIT Technology Review’s editorial staff.

This content was researched, designed, and written by human writers, editors, analysts, and illustrators. This includes the writing of surveys and collection of data for surveys. AI tools that may have been used were limited to secondary production processes that passed thorough human review.

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology.

How AI is changing the economy

There’s a lot at stake when it comes to understanding how AI is changing the economy right now. Should we be pessimistic? Optimistic? Or is the situation too nuanced for that?

Hopefully, we can point you towards some answers. Mat Honan, our editor in chief, will hold a special subscriber-only Roundtables conversation with our editor at large David Rotman, and Richard Waters, Financial Times columnist, exploring what’s happening across different markets. Register here to join us at 1pm ET on Tuesday December 9.

The event is part of the Financial Times and MIT Technology Review “The State of AI” partnership, exploring the global impact of artificial intelligence. Over the past month, we’ve been running discussions between our journalists—sign up here to receive future editions every Monday.

If you’re interested in how AI is affecting the economy, take a look at: 

+ People are worried that AI will take everyone’s jobs. We’ve been here before.

+  What will AI mean for economic inequality? If we’re not careful, we could see widening gaps within countries and between them. Read the full story.

+ Artificial intelligence could put us on the path to a booming economic future, but getting there will take some serious course corrections. Here’s how to fine-tune AI for prosperity.

The AI Hype Index: The people can’t get enough of AI slop

Separating AI reality from hyped-up fiction isn’t always easy. That’s why we’ve created the AI Hype Index—a simple, at-a-glance summary of everything you need to know about the state of the industry. Take a look at this month’s edition of the index here, featuring everything from replacing animal testing with AI to our story on why AGI should be viewed as a conspiracy theory. 

MIT Technology Review Narrated: How to fix the internet

We all know the internet (well, social media) is broken. But it has also provided a haven for marginalized groups and a place for support. It offers information at times of crisis. It can connect you with long-lost friends. It can make you laugh.

That makes it worth fighting for. And yet, fixing online discourse is the definition of a hard problem.

This is our latest story to be turned into a MIT Technology Review Narrated podcast, which we’re publishing each week on Spotify and Apple Podcasts. Just navigate to MIT Technology Review Narrated on either platform, and follow us to get all our new content as it’s released.

The must-reads

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

1 How much AI investment is too much AI investment?
Tech companies hope to learn from beleaguered Intel. (WSJ $)
+ HP is pivoting to AI in the hopes of saving $1 billion a year. (The Guardian)
+ The European Central bank has accused tech investors of FOMO. (FT $)

2 ICE is outsourcing immigrant surveillance to private firms
It’s incentivizing contractors with multi-million dollar rewards. (Wired $)
+ Californian residents have been traumatized by recent raids. (The Guardian)
+ Another effort to track ICE raids was just taken offline. (MIT Technology Review)

3 Poland plans to use drones to defend its rail network from attack
It’s blaming Russia for a recent line explosion. (FT $)
+ This giant microwave may change the future of war. (MIT Technology Review)

4 ChatGPT could eventually have as many subscribers as Spotify
According to erm, OpenAI. (The Information $)

5 Here’s how your phone-checking habits could shape your daily life
You’re probably underestimating just how often you pick it up. (WP $)
+ How to log off. (MIT Technology Review)

6 Chinese drugs are coming
Its drugmakers are on the verge of making more money overseas than at home. (Economist $)

7 Uber is deploying fully driverless robotaxis in an Abu Dhabi island
Roaming 12 square miles of the popular tourist destination. (The Verge)
+ Tesla is hoping to double its robotaxi fleet in Austin next month. (Reuters)

8 Apple is set to become the world’s largest smartphone maker
After more than a decade in Samsung’s shadow. (Bloomberg $)

9 An AI teddy bear that discussed sexual topics is back on sale
But the Teddy Kumma toy is now powered by a different chatbot. (Bloomberg $)
+ AI toys are all the rage in China—and now they’re appearing on shelves in the US too. (MIT Technology Review)

10 How Stranger Things became the ultimate algorithmic TV show
Its creators mashed a load of pop culture references together and created a streaming phenomenon. (NYT $)

Quote of the day

“AI is a very powerful tool—it’s a hammer and that doesn’t mean everything is a nail.”

—Marketing consultant Ryan Bearden explains to the Wall Street Journal why it pays to be discerning when using AI.

One more thing

Are we ready to hand AI agents the keys?

In recent months, a new class of agents has arrived on the scene: ones built using large language models. Any action that can be captured by text—from playing a video game using written commands to running a social media account—is potentially within the purview of this type of system.

LLM agents don’t have much of a track record yet, but to hear CEOs tell it, they will transform the economy—and soon. Despite that, like chatbot LLMs, agents can be chaotic and unpredictable. Here’s what could happen as we try to integrate them into everything.

—Grace Huckins

We can still have nice things

A place for comfort, fun and distraction to brighten up your day. (Got any ideas? Drop me a line or skeet ’em at me.)

+ The entries for this year’s Nature inFocus Photography Awards are fantastic.
+ There’s nothing like a good karaoke sesh.
+ Happy heavenly birthday Tina Turner, who would have turned 86 years old today.
+ Stop the presses—the hotly-contested list of the world’s top 50 vineyards has officially been announced

Separating AI reality from hyped-up fiction isn’t always easy. That’s why we’ve created the AI Hype Index—a simple, at-a-glance summary of everything you need to know about the state of the industry.

Last year, the fantasy author Joanna Maciejewska went viral (if such a thing is still possible on X) with a post saying “I want AI to do my laundry and dishes so that I can do art and writing, not for AI to do my art and writing so that I can do my laundry and dishes.” Clearly, it struck a chord with the disaffected masses.

Regrettably, 18 months after Maciejewska’s post, the entertainment industry insists that machines should make art and artists should do laundry. The streaming platform Disney+ has plans to let its users generate their own content from its intellectual property instead of, y’know, paying humans to make some new Star Wars or Marvel movies.

Elsewhere, it seems AI-generated music is resonating with a depressingly large audience, given that the AI band Breaking Rust has topped Billboard’s Country Digital Song Sales chart. If the people demand AI slop, who are we to deny them?

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology.

What’s next for AlphaFold: A conversation with a Google DeepMind Nobel laureate

In 2017, fresh off a PhD on theoretical chemistry, John Jumper heard rumors that Google DeepMind had moved on from game-playing AI to a secret project to predict the structures of proteins. He applied for a job.

Just three years later, Jumper and CEO Demis Hassabis had led the development of an AI system called AlphaFold 2 that was able to predict the structures of proteins to within the width of an atom, matching lab-level accuracy, and doing it many times faster—returning results in hours instead of months.

Last year, Jumper and Hassabis shared a Nobel Prize in chemistry. Now that the hype has died down, what impact has AlphaFold really had? How are scientists using it? And what’s next? I talked to Jumper (as well as a few other scientists) to find out. Read the full story.

—Will Douglas Heaven

The State of AI: Chatbot companions and the future of our privacy

—Eileen Guo & Melissa Heikkilä

Even if you don’t have an AI friend yourself, you probably know someone who does. A recent study found that one of the top uses of generative AI is companionship: On platforms like Character.AI, Replika, or Meta AI, people can create personalized chatbots to pose as the ideal friend, romantic partner, parent, therapist, or any other persona they can dream up.

Some state governments are taking notice and starting to regulate companion AI. But tellingly, one area the laws fail to address is user privacy. Read the full story.

This is the fourth edition of The State of AI, our subscriber-only collaboration between the Financial Times and MIT Technology Review. Sign up here to receive future editions every Monday.

While subscribers to The Algorithm, our weekly AI newsletter, get access to an extended excerpt, subscribers to the MIT Technology Review are able to read the whole thing on our site.

The must-reads

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

1 Donald Trump has signed an executive order to boost AI innovation 
The “Genesis Mission” will try to speed up the rate of scientific breakthroughs. (Politico)
+ The order directs government science agencies to aggressively embrace AI. (Axios)
+ It’s also being touted as a way to lower energy prices. (CNN)

2 Anthropic’s new AI model is designed to be better at coding
We’ll discover just how much better once Claude Opus 4.5 has been properly put through its paces. (Bloomberg $)
+ It reportedly outscored human candidates in an internal engineering test. (VentureBeat)
+ What is vibe coding, exactly? (MIT Technology Review)

3 The AI boom is keeping India hooked on coal
Leaving little chance of cleaning up Mumbai’s famously deadly pollution. (The Guardian)
+ It’s lethal smog season in New Delhi right now. (CNN)
+ The data center boom in the desert. (MIT Technology Review)

4 Teenagers are losing access to their AI companions
Character.AI is limiting the amount of time underage users can spend interacting with its chatbots. (WSJ $)
+ The majority of the company’s users are young and female. (CNBC)
+ One of OpenAI’s key safety leaders is leaving the company. (Wired $)
+ The looming crackdown on AI companionship. (MIT Technology Review)

5 Weight-loss drugs may be riskier during pregnancy 
Recipients are more likely to deliver babies prematurely. (WP $)
+ The pill version of Ozempic failed to halt Alzheimer’s progression in a trial. (The Guardian)
+ We’re learning more about what weight-loss drugs do to the body. (MIT Technology Review)

6 OpenAI is launching a new “shopping research” tool
All the better to track your consumer spending with. (CNBC)
+ It’s designed for price comparisons and compiling buyer’s guides. (The Information $)
+ The company is clearly aiming for a share of Amazon’s e-commerce pie. (Semafor)

7 LA residents displaced by wildfires are moving into prefab housing
Their new homes are cheap to build and simple to install. (Fast Company $)
+ How AI can help spot wildfires. (MIT Technology Review)

8 Why former Uber drivers are undertaking the world’s toughest driving test
They’re taking the Knowledge—London’s gruelling street test that bypasses GPS. (NYT $)

9 How to spot a fake battery
Great, one more thing to worry about. (IEEE Spectrum)

10 Where is the Trump Mobile?
Almost six months after it was announced, there’s no sign of it. (CNBC)

Quote of the day

“AI is a tsunami that is gonna wipe out everyone. So I’m handing out surfboards.”

—Filmmaker PJ Accetturo, tells Ars Technica why he’s writing a newsletter advising fellow creatives how to pivot to AI tools.

One more thing

The second wave of AI coding is here

Ask people building generative AI what generative AI is good for right now—what they’re really fired up about—and many will tell you: coding.

Everyone from established AI giants to buzzy startups is promising to take coding assistants to the next level. This next generation can prototype, test, and debug code for you. The upshot is that developers could essentially turn into managers, who may spend more time reviewing and correcting code written by a model than writing it.

But there’s more. Many of the people building generative coding assistants think that they could be a fast track to artificial general intelligence, the hypothetical superhuman technology that a number of top firms claim to have in their sights. Read the full story.

—Will Douglas Heaven

We can still have nice things

A place for comfort, fun and distraction to brighten up your day. (Got any ideas? Drop me a line or skeet ’em at me.)

+ If you’re planning a visit to Istanbul here’s hoping you like cats—the city can’t get enough of them.
+ Rest in power reggae icon Jimmy Cliff.
+ Did you know the ancient Egyptians had a pretty accurate way of testing for pregnancy?
+ As our readers in the US start prepping for Thanksgiving, spare a thought for Astoria the lovelorn turkey

For decades, business continuity planning meant preparing for anomalous events like hurricanes, floods, tornadoes, or regional power outages. In anticipation of these rare disasters, IT teams built playbooks, ran annual tests, crossed their fingers, and hoped they’d never have to use them.

In recent years, an even more persistent threat has emerged. Cyber incidents, particularly ransomware, are now more common—and often, more damaging—than physical disasters. In a recent survey of more than 500 CISOs, almost three-quarters (72%) said their organization had dealt with ransomware in the previous year. Earlier in 2025, ransomware attack rates on enterprises reached record highs.

Mark Vaughn, senior director of the virtualization practice at Presidio, has witnessed the trend firsthand. “When I speak at conferences, I’ll ask the room, ‘How many people have been impacted?’ For disaster recovery, you usually get a few hands,” he says. “But a little over a year ago, I asked how many people in the room had been hit by ransomware, and easily two-thirds of the hands went up.”

Download the full article.

This content was produced by Insights, the custom content arm of MIT Technology Review. It was not written by MIT Technology Review’s editorial staff.

This content was researched, designed, and written by human writers, editors, analysts, and illustrators. This includes the writing of surveys and collection of data for surveys. AI tools that may have been used were limited to secondary production processes that passed thorough human review.

Welcome back to The State of AI, a new collaboration between the Financial Times and MIT Technology Review. Every Monday, writers from both publications debate one aspect of the generative AI revolution reshaping global power.

In this week’s conversation MIT Technology Review’s senior reporter for features and investigations, Eileen Guo, and FT tech correspondent Melissa Heikkilä discuss the privacy implications of our new reliance on chatbots.

Eileen Guo writes:

Even if you don’t have an AI friend yourself, you probably know someone who does. A recent study found that one of the top uses of generative AI is companionship: On platforms like Character.AI, Replika, or Meta AI, people can create personalized chatbots to pose as the ideal friend, romantic partner, parent, therapist, or any other persona they can dream up. 

It’s wild how easily people say these relationships can develop. And multiple studies have found that the more conversational and human-like an AI chatbot is, the more likely it is that we’ll trust it and be influenced by it. This can be dangerous, and the chatbots have been accused of pushing some people toward harmful behaviors—including, in a few extreme examples, suicide. 

Some state governments are taking notice and starting to regulate companion AI. New York requires AI companion companies to create safeguards and report expressions of suicidal ideation, and last month California passed a more detailed bill requiring AI companion companies to protect children and other vulnerable groups. 

But tellingly, one area the laws fail to address is user privacy.

This is despite the fact that AI companions, even more so than other types of generative AI, depend on people to share deeply personal information—from their day-to-day-routines, innermost thoughts, and questions they might not feel comfortable asking real people.

After all, the more users tell their AI companions, the better the bots become at keeping them engaged. This is what MIT researchers Robert Mahari and Pat Pataranutaporn called “addictive intelligence” in an op-ed we published last year, warning that the developers of AI companions make “deliberate design choices … to maximize user engagement.” 

Ultimately, this provides AI companies with something incredibly powerful, not to mention lucrative: a treasure trove of conversational data that can be used to further improve their LLMs. Consider how the venture capital firm Andreessen Horowitz explained it in 2023: 

“Apps such as Character.AI, which both control their models and own the end customer relationship, have a tremendous opportunity to  generate market value in the emerging AI value stack. In a world where data is limited, companies that can create a magical data feedback loop by connecting user engagement back into their underlying model to continuously improve their product will be among the biggest winners that emerge from this ecosystem.”

This personal information is also incredibly valuable to marketers and data brokers. Meta recently announced that it will deliver ads through its AI chatbots. And research conducted this year by the security company Surf Shark found that four out of the five AI companion apps it looked at in the Apple App Store were collecting data such as user or device IDs, which can be combined with third-party data to create profiles for targeted ads. (The only one that said it did not collect data for tracking services was Nomi, which told me earlier this year that it would not “censor” chatbots from giving explicit suicide instructions.) 

All of this means that the privacy risks posed by these AI companions are, in a sense, required: They are a feature, not a bug. And we haven’t even talked about the additional security risks presented by the way AI chatbots collect and store so much personal information in one place. 

So, is it possible to have prosocial and privacy-protecting AI companions? That’s an open question. 

What do you think, Melissa, and what is top of mind for you when it comes to privacy risks from AI companions? And do things look any different in Europe? 

Melissa Heikkilä replies:

Thanks, Eileen. I agree with you. If social media was a privacy nightmare, then AI chatbots put the problem on steroids. 

In many ways, an AI chatbot creates what feels like a much more intimate interaction than a Facebook page. The conversations we have are only with our computers, so there is little risk of your uncle or your crush ever seeing what you write. The AI companies building the models, on the other hand, see everything. 

Companies are optimizing their AI models for engagement by designing them to be as human-like as possible. But AI developers have several other ways to keep us hooked. The first is sycophancy, or the tendency for chatbots to be overly agreeable. 

This feature stems from the way the language model behind the chatbots is trained using reinforcement learning. Human data labelers rate the answers generated by the model as either acceptable or not. This teaches the model how to behave. 

Because people generally like answers that are agreeable, such responses are weighted more heavily in training. 

AI companies say they use this technique because it helps models become more helpful. But it creates a perverse incentive. 

After encouraging us to pour our hearts out to chatbots, companies from Meta to OpenAI are now looking to monetize these conversations. OpenAI recently told us it was looking at a number of ways to meet $1 trillion spending pledges, which included advertising and shopping features. 

AI models are already incredibly persuasive. Researchers at the UK’s AI Security Institute have shown that they are far more skilled than humans at persuading people to change their minds on politics, conspiracy theories, and vaccine skepticism. They do this by generating large amounts of relevant evidence and communicating it in an effective and understandable way. 

This feature, paired with their sycophancy and a wealth of personal data, could be a powerful tool for advertisers—one that is more manipulative than anything we have seen before. 

By default, chatbot users are opted in to data collection. Opt-out policies place the onus on users to understand the implications of sharing their information. It’s also unlikely that data already used in training will be removed. 

We are all part of this phenomenon whether we want to be or not. Social media platforms from Instagram to LinkedIn now use our personal data to train generative AI models. 

Companies are sitting on treasure troves that consist of our most intimate thoughts and preferences, and language models are very good at picking up on subtle hints in language that could help advertisers profile us better by inferring our age, location, gender, and income level.

We are being sold the idea of an omniscient AI digital assistant, a superintelligent confidante. In return, however, there is a very real risk that our information is about to be sent to the highest bidder once again.

Eileen responds:

I think the comparison between AI companions and social media is both apt and concerning. 

As Melissa highlighted, the privacy risks presented by AI chatbots aren’t new—they just “put the [privacy] problem on steroids.” AI companions are more intimate and even better optimized for engagement than social media, making it more likely that people will offer up more personal information.

Here in the US, we are far from solving the privacy issues already presented by social networks and the internet’s ad economy, even without the added risks of AI.

And without regulation, the companies themselves are not following privacy best practices either. One recent study found that the major AI models train their LLMs on user chat data by default unless users opt out, while several don’t offer opt-out mechanisms at all.

In an ideal world, the greater risks of companion AI would give more impetus to the privacy fight—but I don’t see any evidence this is happening. 

Further reading 

FT reporters peer under the hood of OpenAI’s five-year business plan as it tries to meet its vast $1 trillion spending pledges. 

Is it really such a problem if AI chatbots tell people what they want to hear? This FT feature asks what’s wrong with sycophancy 

In a recent print issue of MIT Technology Review, Rhiannon Williams spoke to a number of people about the types of relationships they are having with AI chatbots.

Eileen broke the story for MIT Technology Review about a chatbot that was encouraging some users to kill themselves.