Waymo hit 100,000 paid rides per week

Waymo crossed 100,000 paid rides per week.

I want to stay with that number for a moment. One hundred thousand. Per week. Paid. As in, real people, hailing a ride through an app, getting into a car with no driver, arriving at their destination, and paying for the service.

This isn’t a demo. This isn’t a press release about future capability. This is a taxi company. A real, operating, revenue-generating taxi company that happens to have no human drivers.

Alphabet, Waymo’s parent company, reported the number quietly in their earnings call. It wasn’t the headline. It should have been.

The inflection

Every new technology has a moment when it stops being a technology and starts being a business. For personal computers, it was the IBM PC. For smartphones, it was the iPhone. For streaming, it was Netflix hitting 10 million subscribers.

For autonomous vehicles, I think this is that moment.

100,000 rides per week means Waymo is operating at a scale where the edge cases are being encountered and handled routinely. The system isn’t avoiding difficult situations. It’s working through them, thousands of times a day, in real traffic, with real passengers who have somewhere to be.

The safety data supports the scale. Millions of autonomous miles driven. Significantly fewer injury-causing incidents per mile than human drivers. The statistical case for autonomous driving isn’t theoretical anymore. It’s empirical. It’s generated by the same vehicles carrying the same passengers through the same streets, every week.

Why nobody noticed

I think there are two reasons.

First, Waymo doesn’t hype. They don’t have a CEO tweeting about the future of transportation every other day. They publish safety reports. They expand to new cities. They increase ride volume. Quietly. Consistently. Without drama.

Second, the change is invisible if you’re not in a Waymo city. If you live in Phoenix or San Francisco, you see the cars every day. They’re part of the scenery. You might ride in one. For everyone else, autonomous taxis are still an abstract concept, something happening somewhere else.

But 100,000 rides per week means tens of thousands of people are choosing autonomous taxis over human-driven ones. Choosing. Not because it’s novel. Because it works. Because the car shows up, the ride is smooth, and nobody yells at them for the route choice.

What comes next

The question is growth rate. If Waymo maintains the expansion pace they’ve shown in 2024 (adding cities, increasing fleet size in existing cities), 100,000 becomes 200,000 becomes 500,000. At some point, Waymo’s ride volume starts competing with Uber and Lyft in specific markets.

That’s when the economics get interesting. Waymo doesn’t pay drivers. The cost structure is hardware depreciation, fleet maintenance, compute, and insurance. At scale, the per-ride cost should be significantly lower than human-driven ride-sharing.

Lower cost per ride means lower prices for consumers. Lower prices mean more rides. More rides mean more data. More data means better driving. Better driving means fewer accidents. Fewer accidents mean lower insurance. Lower insurance means lower cost per ride.

The flywheel is spinning. It’s still small. But flywheels don’t stay small when the physics work.

The historical marker

I think future historians will look at 2024 the way we look at 1908 (the Model T) or 2007 (the iPhone). Not because the technology was perfected in 2024. But because 2024 is when it stopped being experimental and started being real.

100,000 paid rides per week. No driver. No incidents. In real traffic. In real cities.

The autonomous era started, and most people were looking the other way.

Related thinking: