I’m more skeptical than most that self driving will be properly solved anytime in the next few decades, but I really doubt the article’s claims that it will be able to claim much modeshare from bikes and transit.

Firstly, we already have and have had autonomous vehicles for nearly as long as we have had vehicles, their called taxis and carpools. Making these potentially cheaper, though in practice I doubt it since a taxi’s costs are spread over all its users while a car has to be paid by just you, does not change the fact that they are less convienent than being able to show up and hop on like a bus, or the immunity to traffic delays of rail. Indeed the proposed system of distant out of city parking lots would take more planning than just parking your own vehicle today in most places, as you have to call or order ahead with AVs to have them ready for instead of waking to your car and jumping in. Similarly, getting stuck in traffic does not get much more fun simply because someone else is driving, especially if you can’t even talk to them.

The arguement for them replacing bikes is even worse, because one of the few things proper self driving vehicles are already pretty good at thanks to 360 ultrasonic and lidar sensors at is not blindly running down bikes, and a future with widespread adoption would also imply that most other vehicles have similar driver assistance tech, and as such more people will feel safe biking even in places with shit bike infrastructure. Meanwhile most people who were going to use a bike for a trip will not choose driving over bikeing just because they can get someone else to come pick them up.

I could see it having an effect on modeshare in places with really shit and infrequent transit, but the whole point of rapid transit is that it is more rapid than taking a car. If your transit system is slower and worse than waiting ten minutes in the rain for an Uber, fix your terrible transit system, because that really should be a low bar to clear.