There is probably a pretty well-defined density threshold where vehicle pooling makes sense, vs. personally owned vehicles. In cities, pooled vehicles are a no

There is probably a pretty well-defined density threshold where vehicle pooling makes sense, vs. personally owned vehicles. In cities, pooled vehicles are a no-brainer. It's probably already faster to have an Uber come get you than it is to park a personally-owned car in the core parts of NYC, SF, or some other megacities. This is only going to get better, including specialty vehicles. But as you get out into the suburbs and exurbs, it's more challenging to provide an acceptable level of service (max. 5 minutes from call to arrival, or some other pain threshold that causes people to just wish they had their own car). The economics also start to get harder: if I take an Uber (self-driving or otherwise) out to my house in the sticks, it's a lot less likely to get a return fare back into the city; that trip is a lot less profitable than circulation within an area. So it's doubly bad for a self-driving car company: you need a lot more cars/customer to provide the same service level, and you're spending a lot more time "deadheading" with empty vehicles, or charging extra to compensate. That's not to say that self-driving vehicles won't significantly affect driving and commuting patterns, even in rural areas, but the displacement of personally-owned vehicles is going to be more challenging and may lag urban areas by a very long time.