I'm surprised at how little attention Tesla has given to the idea of robotaxis. At the time they decided getting Mobileye's chips into every Tesla would be a good idea they should have started a dedicated L4 development program, with the big spinning lidars and everything else.

The idea of taking an incremental approach has been abandoned by every serious player except Tesla. Musk is more interested in sizzle than steak. Selling cars and keeping investors excited near term seems a bigger priority than being ready for the day L4 becomes a real thing.

Properly validating a fully autonomous car isn't a sprint, it's a marathon. Tesla is acting like a hare, but I'm betting on the tortoises.