The driverless car finally stopped being a demo and became a service. In the first week of June, Tesla announced that unsupervised Robotaxis are now available across the geofenced Austin metro — a milestone in a race where rival Waymo is already operating paid, fully driverless rides at scale.
Tesla’s Austin push
Tesla’s expansion has been rapid. After integrating unsupervised vehicles in a limited way in January, the company widened the program to Dallas and Houston in April and has now blanketed the Austin metro with robotaxis that operate without a safety driver. For Elon Musk, who has staked enormous credibility on autonomy, removing the human from the seat across an entire metro is the proof point skeptics demanded.
Waymo’s lead
Waymo, meanwhile, is the operator to beat. It runs revenue-generating robotaxi service in Phoenix, San Francisco, Los Angeles and Austin, with what analysts call empirically defensible safety performance and active plans to add cities. Waymo’s methodical, city-by-city rollout has built a meaningful lead — the gap to other operators, by most measures, is real.
Two philosophies
The contrast in approach is the story. Tesla bets on a vision-only architecture — cameras and neural networks, no lidar — aiming for general autonomy that could one day run anywhere, and ties it to Cybercab production and consumer Full Self-Driving. Waymo uses a richer sensor suite of lidar, radar and cameras plus high-definition maps, trading flexibility for reliability inside defined operating zones. It is scale-with-guardrails versus bet-on-general-AI.
Why it matters
Driverless ride-hailing at metro scale reshapes transportation economics, urban planning and millions of driving jobs. It also raises the stakes on safety: every incident is scrutinized, and public trust is fragile. The companies that prove they can operate safely and cheaply at scale stand to define a multi-hundred-billion-dollar market.
The skeptic’s view
Caution is warranted. Geofenced ‘available’ is not the same as ubiquitous, and Tesla’s vision-only approach still faces hard questions in bad weather and edge cases. Waymo’s expansion is real but deliberate. The technology works in defined conditions; generalizing it everywhere remains the unsolved problem.
The bottom line
With Tesla going driverless across Austin and Waymo scaling across four cities, 2026 is the year robotaxis became a real, paid service rather than a promise. The remaining question is no longer whether driverless works, but how fast — and how safely — it can spread.
Photo: Wesley Fryer / BY-SA via flickr