One car owned per adult is very inconvenient to deviate from in a car friendly urban environment, even if robotaxis can be made to have overall cost parity to car ownership.
The net result, I think, is that robotaxis can only ever increase the total number of cars and concurrent drivers.
Would you disagree? Is there some compelling reason I'm missing the motivation for families to switch from 2.5 cars to 1 because of robotaxis? Or, similarly, for single people to switch from 1 car or no car, just transit, to robotaxis?
Robotaxis seem to me to have potential well solve the niches human taxis are bad at: late night safe driving, very long point to point going out of an urban core into a rural area, perhaps even personal cargo shipping, if a robo van could enable you to do furniture shopping without renting a truck. But these are edge cases that compete with private cars pretty intensely. I'd only prefer a late night taxi if I'm drunk. I'd only want a robotaxi to a rural area if I'm an urban living rock climber or hiker, and I don't care to own a car. The total available market is just wildly incompatible with the current and speculated valuation, as far as I see it.