Robotaxis in San Francisco
Back from San Francisco. Let’s talk robotaxis.
First, the number of them. Way more than I expected.
Product angle: Waymo charges 2x to 5x the Uber price for the same trip, and still wins the ride. Consumers value novelty. But after a few conversations, the real driver is having no stranger in the car1.
As a passenger, the experience is excellent. Waymo, Zoox (it’s not a car), Cybercab (with a human onboard). The app flow is clean: enter destination, point the phone at the car, press start. Off you go.
I watched confrontations in the Tenderloin between people and vehicles. Sometimes funny, often sad. Human drivers show zero mercy to Waymos, which already struggle to change lanes. Will the algorithm eventually be allowed to get more aggressive on merges?
The right vehicle shape is still up for debate. But the strategy that wins is flooding the streets. Models without a steering wheel will get there first: lighter, fewer recharges.
A specific tax (TCM) applies to autonomous vehicles and Uber/Lyft rides. A few percent on top of every fare.
Far fewer cyclists than in Paris, though plenty of new bike lanes have appeared since my 2018 trip. Waymos are attentive. Door-opening alerts. And they don’t pull back into the lane when a cyclist passes on the right. The upside of 29 eyes instead of 2.
Passengers told me Uber/Lyft drivers in SF often come from out of town. They don’t know the city any better than a Waymo does. And women passengers were very clear about their personal safety expectations, after several incidents involving drivers. ↩︎