Tesla’s robotaxis are moving in reverse

Tesla's robotaxi program is running into the kind of arithmetic that does not care about press releases. Paid miles traveled fell 36% in the second quarter, even as the service expanded into new cities. More geography, fewer rides. That is the headline.

The contradiction is the story. Expansion is supposed to produce growth. When it produces the opposite, the explanation usually lives in one of three places: pricing, supply, or demand. Tesla's own figures do not clarify which, but the direction is unambiguous. The unit economics of the program are not improving with scale. They are deteriorating.

For a company that has tied a significant portion of its future valuation to autonomy, the optics matter. Robotaxi miles are the most direct, verifiable proxy for whether the technology is generating real revenue in real conditions. A 36% decline in a quarter is not a rounding error. It is a signal that something in the operating model is not holding together as the footprint grows.

The remote-work angle here is incidental but worth noting. Robotaxi demand is shaped by commuting patterns, urban density, and the rhythms of people leaving their homes to go somewhere. If those patterns are softer than the pre-launch models assumed, the fleet is carrying capacity it cannot monetize. The technology may be ready. The market it was built for may not be.

Tesla has time, capital, and a long history of turning operational stumbles into narrative comebacks. But the numbers now belong to the public, filed in the same quarter the company expanded service. That juxtaposition is harder to spin than a single bad data point would be. The road ahead is longer, and the miles are shorter, than the map suggests.