If you assume that driverless cars will work, this is a neat perspective:
If you live in a big city, go outside and look up and down the street. You can see hundreds of autos just sitting there doing nothing. Some of these cars move less often than once per week. I don’t live in a big city and my car moves 2 hours a day on a day of heavy use.
Now envision 5-10 years from now…you go out on the street and there are no cars…because there’s an Enterprise rent-a-car parking garage with hundreds of self driving cars two miles from where you live. They have a smartphone app that you press a button for what kind of vehicle you need and it shows up out front in five minutes.
There’s more here on the efficiency gains when humans don’t need to operate cars (we’ll need less of them). And trucks, too.
Not Unreasonable 
Interesting to think about, but we already have the equivalent of driverless cars. They are called taxicabs. You dial a phone number, and a car arrives. It will take you wherever you want to go.
Driverless cars are likely to replace taxi drivers, but they will only replace family-owned autos to the degree that the marginal cost of a driverless car is less than the marginal cost of a taxi. The shape of the demand curve comes into play, too.
And it’ll still be tough to get a car on Thursday night in Midtown.
Good point. There’s something about owning your own transportation which is not going away.
There’s some speculation about increased utilization of trucking fleets at the link, too, which I like a bit more. The real constraint there is driver fatigue.