Riding in three Waymo taxis in San Francisco finds two avoided traffic while another didn't, the app works like Uber, and the first experience can be confusing
On Monday, Waymo began letting the public pay for rides in its driverless cars in San Francisco.
Context & Ripple Effects
Waymo's paid public service in San Francisco builds on its earlier permit to charge riders in the city, moving the product from supervised ride-hailing toward a consumer-facing driverless service.
The rider account captures the adoption challenge at this stage: booking can resemble a familiar ride-hail workflow, but the vehicle's routing choices and the unfamiliar first ride still shape trust. Later coverage of Waymo removing its San Francisco waitlist suggests that access, not just autonomy, became the next scaling lever.
First-order effects
- San Francisco riders can pay for a driverless Waymo trip, giving Waymo real customer experience data beyond testing and limited access.
- Waymo's immediate product burden extends beyond safe driving: it must make pickup, routing behavior and first-time rider guidance intelligible enough for paying customers.
Second-order effects
- Uber-like app mechanics lower the switching cost for riders, so service reliability and trip experience—not booking familiarity—become the clearer point of comparison with conventional ride-hailing.
- Inconsistent apparent traffic avoidance can make routing transparency and customer support more important as Waymo seeks repeat use rather than novelty rides.
Third-order effects
- Robotaxi competition is likely to be decided increasingly by operational consistency and consumer onboarding alongside autonomous-driving capability; early riders’ tolerance for confusing edge cases may not hold at broader scale.
- If public access expands, cities and operators will face a more visible standard for how driverless services explain vehicle decisions and handle rider uncertainty.
The trend: Robotaxis are moving from permit-led pilots toward consumer mobility products, where repeatable service experience becomes as consequential as the autonomy stack.