Operating a incredibly limited test program while burning 100s of millions is not my definition of solving general autonomy.

Maybe if they role out this program to city after city in short intervals and have 10000+ taxi operating I would consider it.

Waymo does some things Tesla doesn't do, Tesla does things Waymo doesn't do. Tesla clearly has no interest in offering a money losing taxi service in one small city. Just as Waymo has no interest in doing all the work to figure out the meaning of obscure signs on roads in China.

The question is, when can I click on an app, get a car and drive basically anywhere. That is 'the solution'.

> driverless paid service

Which works in one small limited area of the earth because of high-def maps, which is not what you would normally call "scalable".

At some point there was also a car company that only sold a single electrical car model, it was pretty expensive and they didn't make that many cars per year, which is not what you would normally call "scalable."

Back to the original goalposts, as far as I know, Waymo is the only company that at this point, does offer a commercial service that's fully driverless. As you point out, they don't cover all of Earth, but neither do any of their competitors. Even Uber and Lyft, with human drivers, don't support all pick up and drop off locations, they only work in certain service areas.

Yeah, that whole car's tech didn't rely on expensive highdef maps.

> fully driverless

Don't they have safety drivers that closely monitor through wireless connection and often take over? While it might look driverless, it is not technically. This is also another reason why it doesn't scale.

Right, because who would digitalize the entire world and take pictures of every house and store front - that'd be insane!