All-New Waymo Robotaxi Finally Debuts

The four-year development cycle is the detail worth sitting with: Waymo spent that time not just refining perception software but rebuilding the vehicle as an integrated system, which means competitors can't simply license the software stack and catch up. The hardware and software are now co-designed in ways that make replication slower and more capital-intensive than the prior generation.

The data acquisition angle connects directly to the Verge story from the same day about Shift filming users doing household chores to train embodied AI. Waymo's production fleet is, among other things, a proprietary data flywheel: every mile driven generates labeled real-world sensor data that competitors cannot replicate from simulation alone. The Shift story illustrated how data scarcity has become the binding constraint for embodied AI broadly, and Waymo's scaled deployment is the autonomous vehicle equivalent of that same dynamic, except Waymo controls both the collection infrastructure and the downstream model training.

Watch whether Waymo announces a fleet expansion target or a new city launch within the next six months. If they do, it confirms the production hardware is genuinely ready for scale rather than a limited showcase rollout.