Waymo: Already Running What Tesla Is PromisingThin moat

Tesla's valuation rests substantially on autonomous driving, and the honest assessment of whether it will arrive is a root threat on this company. This page is narrower: on the fact that a competitor is already operating a driverless commercial service, and did it by rejecting every technical choice Tesla made.

Waymo built its system on detailed prior mapping, lidar and radar alongside cameras, and geographic caution — expanding city by city, each one surveyed in advance. Tesla argued that approach could never scale: too expensive, too slow, too dependent on maps that go stale. It bet instead on cameras alone and general-purpose learning, which if it works produces a system that drives anywhere rather than in mapped territory.

The uncomfortable position is that the approach Tesla called unscalable is carrying paying passengers at meaningful volume today, while the approach Tesla called inevitable is still expanding under supervision. Being right eventually is worth a great deal in this industry; being right late is worth much less, because the incumbent operator accumulates regulatory approvals, safety data and city relationships that a later entrant must acquire separately.

Watch driverless miles without a safety driver, by company. That single metric cuts through every claim made about autonomy, and it is the one on which Tesla must eventually close a gap rather than argue about architecture — with an automotive gross margin of 16.9%¹ funding the attempt.