I think taxis have a place even in cities with great transportation. I live in Toronto and 90% of my commuting is walking, 8% public transit and 2% driving. But there are some trips that would be very difficult to do a way other than driving (for example carrying lots of stuff or awkward cargo) and taxis fill that gap wonderfully. Especially if self-driving taxis could handle long trips a lot better as inter-city is a place where Toronto public transit unfortunately sucks (for example visiting my parents in cottage country).
Mass transit only works when there are masses of people, and most of the US doesn't have that. But places like the Great Plains and the rural Northeast already have comprehensive roads systems, and electric robot cars turn those roads into transit for people who, currently, must have cars.
In rural places, not only are there no buses or trains, there's hardly any taxis. Maybe during the day if you wait an hour or two, but you're not getting home at night without your own car.
A couple robot taxis roaming around every rural country in the US comprehensively solves this problem.
Agree. Though I imagine the cost to run robot taxis in low density areas would be much higher than in high density areas, as there'd be a higher distance between rides, and so more wasted time and gas. I suspect that would slow adoption.
Also if we dream big here, I wonder how robot buses would work in low density areas, maybe with dynamic routes and pricing?
This is true for intersection throughput, but I bet full travel throughput (walking to the bus, waiting for the bus/train, walking to your destination) is the same or better with Waymo.
Would it be way better to make walkable neighborhoods, mixed-use developments, and reliable and frequent public transit?
Yes. Yes it would.
But, in lieu of that, self-driving has a lot of advantages in the long run, even if the technology isn't 100% perfect right now.