I work in the land development industry. By far, the biggest limiting factor in developing parcels of land in California is the space available for parking. Whether it be housing, office, or retail space, the need to park drives most site plans and thereby the underlying economics.

Driverless cars have the potential to disrupt parking requirements in a profound way.

Since drivers will no longer need to park near where they work, their car can drop them off and go to a remote location to park until needed.

Related to the first point, there really isn't any need for a car to be parked at all. Many commuters will put their cars into a Uber-type system where others can call up their car for taxi travel while the owner is working at their desk. People will make money on the side by allowing their cars to act as taxis when not in use. Or course, some people won't want strangers in their cars, but many will, and the vast parking lots of unused cars currently surrounding nearly every workplace will no longer be necessary.

One of the early problems will be congestion from people who try to avoid parking tolls by just sending their car on random drives to avoid parking altogether. The streets will become defacto parking lots as cars roam aimlessly until their owners need them.

Without the need to store automobiles in or near downtown areas, urban cores will blossom. Development densities can be greatly increased as we will be able to more easily house and work more people as the space-hungry automobiles are pushed out.

In short, self-driving cars will spawn a revolution in land use, and it will help solve some of our more intractable problems with development density and cost.