Waymo: 'Cameras Are Incredible, But They Aren't Enough'

As Tesla prepares to launch its Cybercab — a robotaxi with no steering wheel or pedals — Waymo has issued a pointed warning about Elon Musk's approach to autonomous driving, without mentioning either by name.

Srikanth Thirumalai, Waymo's VP of onboard software, published a post titled "10 AI Lessons from Driving 200+ Million Fully Autonomous Miles." The intended reader is not hard to identify.

The claim

"Cameras are incredible, but they aren't enough," Thirumalai writes. "For years, there's been a debate over whether cameras alone could solve full autonomy. Now, after more than 200 million real-world miles, the data is clear: safe, fully autonomous operations at scale require more."

Waymo pairs cameras with lidar and radar to build what he calls "a rich, redundant world view that no single sensor can replicate."

The bet each company made

The disagreement is old and expensive. Tesla removed lidar on the argument that human drivers manage with two eyes, so a sufficiently good vision model should too — and that the sensor's cost made a mass-market robotaxi impossible.

Waymo kept the redundancy on the argument that cameras fail in correlated ways, and that a system with no independent check on its own perception has no way to know when it is wrong.

The timing of the post is the argument. Waymo is staking its position on the record before Tesla's vehicles start carrying passengers without controls, so that whatever happens next can be read against a claim someone made in advance.