> not to the point where anyone would realistically design a city center around it

Sure. Neither is Phoenix's light-rail system, for the most part. These things take time to play out and gain buy-in.

Americans take about 34 million public-transit trips a day [1]. Assuming 25 rides per day, that's about 1.4 million self-driving cars to rival public transport's impact. Waymo has "about 3,000 robotaxis deployed nationwide" [2]. Doubling fleet size annually–Waymos and non-Waymos, though currently they have no peers–would get us to parity in less than 10 years. (A more-realistic 35% growth rate puts us around 20 years.)

The point of that excercise is to say that within 10 to 20 years, less time e.g. California's HSR or New York's Second Avenue Subway took to get online, we could see as many trips in AVs in America as we do on public transit of all types. That's close enough to start looking ahead to.

[1] https://www.apta.com/news-research/about-the-industry/public...

[2] https://www.axios.com/local/san-diego/2026/03/30/waymo-speed...