Could this disruptive technology force a reckoning with the role of cars in cities?
More likely IMO that regulations fall on robotaxis while thoroughly exempting "traditionally" driven cars giving them a similar privilege to single-family detached owner-occupied homes.
I'm not so sure how to take this piece (and its mate in Technology Review). E.g. if the question is "What problem do robotaxis solve?" the obvious answers are "Vehicle is reused during the day and can be parked outside of the central city," "User doesn't have to pay the capital costs of owning a vehicle," and "Doesn't require a driver." The latter point shows that Uber/Lyft drivers will be put out of work by technology, but that's been happening from before the Luddites through the elimination of the taxi industry, and the public's position has always been "They'll have to get new jobs." The public has always favored lower prices for something they feel they must consume.
There's worry that robotaxis won't be adequately regulated, but it seems like they're *already* regulated by the Calif. Public Utilities Commission and presumably any general objection on the part of the voters of San Francisco will be heard. And certainly whoever is insuring the robotaxis is paying close attention to their safety record. (Though there seem to have been only 1 million vehicle miles logged so far, whereas US highway fatality rates are in the 1-per-100 million vehicle mile range, so we're not getting much useful data yet.)
If the question is the pragmatic politics of creating a city where autos aren't a major form of transportation, it seems to me that the first question isn't so much how to get there but what the destination is. I've seen a few European city centers that are nicely walkable, but they seem to be very "housing tight", either extremely expensive or extremely rent-controlled; in either case, someone of modest means can't expect to move into it. (Paris' outlying suburbs being a notorious consequence.)
A problem is the metric "How many jobs are within tolerable commuting distance (e.g. 30-45 minutes) from where I live?" I'm a computer programmer, which is a fairly typical job these days, and I need to have a choice among at least 50 similar employers to ensure I see a good job market. The road system of Boston more or less replicates that for several million people of several hundred different occupations, but it would be tough to do with any sort of mass transit, most of which are at best hub-and-spoke services for dense financial districts.
A less talked-about problem is the segregation of public education. Currently, the suburbs and exurbs provide "good schools", that is, they use snob zoning to keep poor kids out of the schools, and relatively affluent parents pay dearly for that service. (Immediately after their children graduate from high school, they sell those houses and buy ones in cheaper/less affluent areas.) As long as public schooling is organized by geographic catchment areas, there's going to be demand for geographically dispersed housing.