Tesla touts 380,000 unsupervised robotaxi miles with ‘zero notable incidents’

Tesla's robotaxi fleet has reportedly logged 380,000 unsupervised miles across six cities in two states, with the company's AI chief declaring "zero notable incidents" on the second-quarter earnings call. The figure is being framed as vindication of Tesla's camera-only approach to autonomy, a deliberate counterpoint to the lidar-and-radar stack used by Waymo and most other serious players in the space.

The number itself is modest by autonomous-driving standards. Waymo has been operating commercial driverless service since 2020 and now runs tens of millions of fully driverless miles per quarter. Tesla's 380,000 cumulative figure is a rounding error against that denominator. What matters for investors is the growth curve: Tesla says unsupervised miles are compounding at a double-digit weekly rate, and management expects that pace to continue through year-end. If that trajectory holds, the fleet's exposure expands by roughly an order of magnitude every few months, which is the only metric that will eventually make the safety claim statistically meaningful.

The "zero notable incidents" language deserves scrutiny. Tesla's own definition, as stated on the call, excludes any event where another road user struck a stationary Tesla vehicle. That carve-out is significant. A robotaxi stopped at an intersection that gets rear-ended, or one that fails to clear an intersection and is hit by cross traffic, would not count under this framing. The metric measures incidents initiated by the Tesla vehicle, not incidents involving one. For a service that is supposed to replace human drivers, the distinction matters: a human driver is also responsible for positioning the vehicle safely, not just for moving it.

The strategic subtext is the camera-versus-lidar debate. Ashok Elluswamy's remarks were aimed squarely at the "experts" who have argued that cameras alone cannot deliver safe autonomy in adverse weather, low light, or visually ambiguous conditions. Tesla is betting that scale and data will close whatever perception gap exists, and that the cost advantage of a camera-only stack will let it deploy faster and cheaper than lidar-equipped competitors. The earnings call is the venue where that thesis gets priced in, and the 380,000-mile figure is the supporting evidence management chose to lead with.

For the labor market, the implications are still largely theoretical. Tesla's robotaxi service is not yet a mass-market product, and the cities in operation remain limited. The more immediate workforce story is the construction and staffing of the chip manufacturing facility Musk announced separately, and the ongoing Senate testimony about regulating the industry. Remote and distributed-work dynamics are not materially present in this story; the connection to hiring and compensation is indirect at best, and the source does not support a labor-market angle beyond noting that autonomous-vehicle deployment will eventually reshape driving as an occupation.

The near-term catalysts to watch are straightforward: whether the weekly compounding rate actually persists, whether Tesla expands into additional states rather than continuing city-by-city, and whether regulators accept the "zero notable incidents" framing or impose their own reporting requirements. The Senate committee hearing on autonomous-vehicle regulation, referenced in the coverage, is the venue where that definitional fight will likely play out.