I see a number of challenges with the assumptions:

1) Tesla is no longer the world's leader in EV's or in autonomous software and its market share will likely erode in the future.

2) Major cities around the world frequently have robust mass-transit systems and, even with the current availability oof manned taxis and car-share, locals find it more cost-effective to take the bus/subway. While this is likely similar in cities like Paris and Zurich, I'll address NYC. With a nighttime population of 8.5 million and about 12 million during the day, and with a relatively affluent population with over 400,000 millionaires (despite rumors to the contrary, showing a large post-COVID increase in residents), New York City has only 13,587 yellow medallion taxicabs (though it has quite a few car service and ride-share cars).

3) In the world's most populous, but poor, environments (e.g.: India), taxis are so cheap that even poor locals use them instead of purchasing cars, etc. While there is a "taxi culture", there is no profit in the system. The hierarchy is Air-Conditioned cab, taxi without A/C, tuk-tuk, motorcycle taxi (depending on number of people traveling and what they are lugging with them). It's hard to see how autonomous taxis can compete with drivers earning $2-3 a day.

4) In affluent environments which are lacking good mass transit, such as the rural and suburban US, there is a perceived value to owning a car as goods and services tend not to be located close to housing in rural areas.

Unless and until the US and European population is willing to give up car ownership wholesale or the competition in third-world countries becomes much more expensive, auto-taxis may continue to be a pipe-dream in many areas as a profitable business.

Jeff