Autonomous vehicles have a lot of potential. As long as you program them right, they won't speed, won't break traffic laws, and won't get drunk, high, abusive, or violent. And the technology has been getting much more capable, even as some of the hype has died down, taking some of the related companies with it. Waymo still easily leads the field and is already operating commercially in six cities across America, with a dozen more (plus London) coming soon. Waymos can even drop you off and pick you up at the airport in Phoenix and San Francisco.

Soon, Waymo will begin deploying its sixth-generation Waymo Driver, using upfitted Zeekr Ojai minivans, adding to the Jaguar I-Paces that have become so common on San Francisco streets and to its fleet of Hyundai Ioniq 5 electric vehicles. It has upgraded the cameras, lidar, and radar, meaning the cars can better sense their environments at night and in inclement weather. There are even microphones that can pick up sounds like sirens to better inform the robotaxi of the direction the emergency vehicle(s) are coming from.

But even with all these advances since the pod-like two-seater that predates even the Waymo name, there are still a few things that remain beyond a robotaxi's capabilities. Like closing a door a passenger left open on their way out. All the sophisticated sensors and high-powered computer processing in the world are useless if the car can't move until the door closes and there's no one there to give it a hand.

Read full article

Comments

© Waymo

Google-spinoff Waymo is in the midst of expanding its self-driving car fleet into new regions. Waymo touts more than 200 million miles of driving that informs how the vehicles navigate roads, but the company's AI has also driven billions of miles virtually, and there's a lot more to come with the new Waymo World Model. Based on Google DeepMind's Genie 3, Waymo says the model can create "hyper-realistic" simulated environments that train the AI on situations that are rarely (or never) encountered in real life—like snow on the Golden Gate Bridge.

Until recently, the autonomous driving industry relied entirely on training data collected from real cars and real situations. That means rare, potentially dangerous events are not well represented in training data. The Waymo World Model aims to address that by allowing engineers to create simulations with simple prompts and driving inputs.

Google revealed Genie 3 last year, positioning it as a significant upgrade over other world models by virtue of its long-horizon memory. In Google's world model, you can wander away from a given object, and when you look back, the model will still "remember" how that object is supposed to look. In earlier attempts at world models, the simulation would lose that context almost immediately. With Genie 3, the model can remember details for several minutes.

Read full article

Comments

© Ryan Whitwam

The self-driving car company, an arm of Google’s parent company, is rapidly expanding its service beyond a handful of cities.

© Caitlin O'Hara for The New York Times

The automaker also said it would invest $2 billion in xAI, the artificial intelligence company controlled by its C.E.O., Elon Musk, and stop making the two oldest models in its lineup.

© Annie Flanagan for The New York Times

From talking computers to self-driving cars, here are the trends to watch.

© Sisi Yu

Shares of Tesla have hit new highs on optimism about the company’s self-driving taxis. But experts say Tesla is far behind Waymo, which has a big head start.

© Ariana Gomez for The New York Times

The self-driving cars came to a halt at intersections when the power outage knocked out traffic signals, causing tie-ups but no accidents or injuries.

© Tayfun Coskun/Anadolu, via Getty