The Frontier

Waymo now provides over 500,000 fully autonomous rides each week across 11 US cities with a fleet of roughly 3,000 vehicles.

Dmitri Dolgov says Waymo has moved past core technology development into a phase of accelerated global scaling. Dolgov foresees any major metropolitan area eventually having Waymo service.

Waymo's technical architecture is built around a large off-board AI foundation model that understands physical world dynamics and social driving norms. This model is specialized into three core teachers: the driver, the simulator, and the critic.

Waymo uses a multi-sensor fusion stack of cameras, LiDAR, and radar. LiDAR provides high-resolution 3D maps, radar excels in adverse weather, and cameras offer rich visual data. The encoder-decoder AI architecture uses intermediate world representations between sensing and action.

The sixth-generation Waymo vehicle features a custom passenger-centric design with sliding doors and more interior space. Dolgov says its new sensor suite is simpler, more capable, and a fraction of the cost of previous generations.

Dolgov says Waymo's upcoming international expansion into London and Tokyo this year still requires data collection and specialization, but the core driving technology generalizes well. Cold winter weather presents a hardware and software challenge for the full stack.

Global VC investment in AI hit a record $242 billion in Q1 2026, with 64% focused on OpenAI, Anthropic, XAI, and Waymo. This represents roughly $3 billion invested per day.