Executive Summary
Waymo delivers 500,000 fully autonomous rides weekly across 11 US cities, marking a decisive shift from research to commercial scaling. Co-CEO Dmitri Dolgov reveals the technical architecture behind this breakthrough: a foundation model that specializes into three core systems—the driver, simulator, and critic—then distills into efficient on-vehicle inference. This approach solved the fundamental challenge of achieving superhuman safety while maintaining operational efficiency. The sixth-generation platform reduces sensor costs to consumer ADAS levels while expanding to custom vehicles and international markets including London and Tokyo this year. Dolgov explicitly states 'we've clearly moved past the stage of scientific research and deep core technology development to this new phase of accelerated global scaling.' The technical moat stems from closed-loop reinforcement learning requiring realistic simulation and critic models—capabilities that pure end-to-end approaches cannot replicate at scale. With Google's patient capital enabling 17 years of development, Waymo now represents the clearest validation of autonomous vehicle commercialization, positioning Alphabet to capture significant value from the transportation automation market while demonstrating the strategic value of sustained technical investment in emerging technologies.
Key Insights
what Dmitri Dolgov said“You start with the foundation model. Then you specialize in fine-tune, still-off-board model, those are the teachers, and then you distill, each one of the teachers kind of distill, trains its own student and the driver, the simulator, the critic.”
This is a preview. Log in to see the full analysis including investment opportunities, risks, catalysts, and detailed insights.