Why Tesla's 20 Robotaxis Are Worth More Than Waymo's 3,500
By the scoreboard the news keeps showing you, Tesla is losing the robotaxi race, and badly. About twenty driverless cars carry paying riders, in one city, while rivals run thousands. But that scoreboard is measuring the wrong thing. Because Tesla also sits on well over a million cars, already sold and on the road, that could become robotaxis with a single software update sent overnight.
This is not a race to build cars. It is a race to switch cars on, and then to prove they are safe with nobody watching. So this video gives an honest picture of where Tesla's robotaxi scaling actually stands in the summer of 2026: what has really launched, what is still just a promise, and the handful of things over the next year that decide whether this stays a trickle or turns into a flood. No hype, just evidence. And the cautious, slow-crawl case gets the weight it deserves.
In this episode
- What Tesla has actually launched vs what is still a promise (~20 driverless cars; paid rides in Austin, Dallas, Houston, Miami; still supervised by remote operators)
- The two ways Tesla adds robotaxis, and why one moves at the speed of a factory and the other at the speed of a download
- Why every paid ride today still runs on an ordinary Model Y, not the Cybercab (100+ built, zero carrying a rider)
- The honest Waymo reality check: 577 vs 42 in Texas, ~500,000 paid rides a week, ~$75,000 a car
- The four milestones that decide the next year: the owner network, unsupervised driving, Cybercab in service, and new states
- The one number the whole ramp hinges on: the switch-on rate