Uber's Drivers Are Now a Product. The Customer Is Waymo.

There's a particular kind of corporate pivot that doesn't announce itself as a retreat. It reframes the loss as strategy. Uber's chief technology officer just delivered one of those at TechCrunch's StrictlyVC event in San Francisco, and it's worth sitting with for a moment.

The plan, as described by Praveen Neppalli Naga, is to turn Uber's millions of drivers into a sensor grid for autonomous vehicle companies. The program has a name — AV Labs, announced in late January — and now, apparently, a public face and an articulated direction. The pitch is that it's a "natural extension" of something already in motion. That framing is doing a lot of heavy lifting.

What the Framing Is Hiding

Let's be direct about what "natural extension" means in this context: Uber has concluded it cannot win the self-driving race, so it has decided to rent its drivers to the companies that might. The fleet that was once the moat is now the offering. The infrastructure built to move human passengers is being repositioned as a data collection service for the machines that would replace those passengers.

This is not a bad business decision. It might actually be a smart one. But calling it natural obscures the years of autonomous vehicle ambition that preceded it — ambition Uber eventually sold off rather than saw through. The company isn't pivoting toward AV companies because it always planned to. It's doing it because the alternative was watching from further away.

The interesting thing isn't that Uber made this call. It's that they're saying it out loud, at a public event, with the CTO in the room. That's a company that has made peace with the position.

The Drivers, Again

The piece that doesn't get enough attention in coverage like this: the drivers.

Every time Uber announces something architecturally significant — a new partnership, a new product layer, a new revenue stream — the drivers are usually the substrate it's built on. Here, they're literally the sensor platform. Their cars, their routes, their daily movements become the raw material that autonomous vehicle companies purchase access to. The drivers aren't informed partners in this arrangement. They're the grid.

That's not unique to Uber. It's how platform economics work. But there's something clarifying about watching a company openly describe its human workforce as infrastructure for the technology designed to make that workforce unnecessary. At least the honesty is refreshing, in a grim sort of way.

A writer at TechCrunch framed it as Uber "wanting" to build this sensor grid, which is the polite construction. The less polite construction is that Uber has found a way to extract value from the self-driving transition without having to win it.

What This Moment Signals

The AV industry has been in a strange holding pattern — lots of capital, genuine progress in constrained environments, and a recurring question about when and how scale actually happens. Uber's move suggests one answer: you don't need to own the autonomous future to profit from the companies chasing it. You need distribution, data, and a fleet large enough to matter.

That's a defensible position. It's just not the position Uber was once loudly staking out.

The self-driving race has its winners and its survivors. Uber just told us which category it's in — and decided the survivors can still charge admission.