A self-driving car that sits idle is a wasted resource, so the owner will always want to put it to some profitable use by moving people or goods. Contrast that to human-owned vehicles that spend 95% their life parked: https://www.reinventingparking.org/2013/02/cars-are-parked-9... That's a tremendous amount of the earths mineral resources that were mined, processed, and transported (at great carbon expense) to sit around doing nothing.

> A self-driving car that sits idle is a wasted resource

This is also true of non-self driving cars, yet we structure our resources around this case. Car use has extreme & predictable peaks with long down periods between them. If we couldn't prevent it the first time what are we doing differently to prevent it the second time?

And again rideshare has given us a slight preview of this. The taxi constraints incentivize having cars near their predicted users. We'll either need to build and maintain waiting-car storage in central areas (eg parking), or accept idle circulation in down times.

Who is going to use these cars when everyone is in school or at work or sleeping. This will not solve any issues since most people follow the same schedule.

Transportation demand would still be "lumpy", but surge pricing would encourage people to shift their transportation needs to save money. Deliveries would probably be one of the common uses during the workday. Utilization would not be 100%, but it's certain to be more than the 0% productive use that parked cars currently have.