Then you should be hoping for more technology not less. Imagine the day Google automated cars signs a deal with Halo or Uber. The utilization of most cars is under 10% anyway - and as they are able to consolidate information, they can even help to provide "carpooling" options from door from the suburbs to even further reduce the cost of transportation and car ownership.
Automated cars also mean much greater efficiencies on roads and highways (gone are the days of long pileups and people who don't know how to merge and people who drive with their brake lights always on).
And now with Automated parking garages... this means cars will be able to be parked much closer to where people will use them (e.g. downtown). It would be even more interesting if existing garages can be converted so that they can carry even more capacity - and it would also save a ton of time of having to drive say 3-4 flights from your space in the garage. Add this all up and we're going to hit "peak cars" soon enough - and after this tipping point, the number of cars on the road could actually fall quite dramatically. And we're not even talking about the substantial improvements that are being made to battery technology + even engines (given how inefficient internal combustion engines are - ie well under 20%)
Versions of these exist throughout the world and they are successful at what they do. Sort cars in to as most space efficient of a space as possible. I'm reluctant to continue though with what you originally post ... that there's a positive network effect for this with automated driving systems.
Parking is basically useless space when you consider it from a land use perspective. The public would gain more benefit if we didn't need as much parking as we have. The sad thing is that after a certain point, the more parking we have, the more parking we need as this starts to overflow its effects into mode shift away from more space and socially efficient modes of transport--walking, biking, public transit, shared vehicles.
I guess my premise here is best reserved for another conversation, another forum post. But, greatest efficiency is not found in semi-personal or personal automobiles. Efficiency as measured by volumes of vehicles, yes. But, efficiency as measured by number of people. No.
These technical, engineering marvels are awesome, but they will not get at the major problems facing society -- they are marginal changes, not the fundamental changes that we need.
As to your assertion that self-driving cars + automated parking garages + car-sharing will magically deliver us to a post-peak car situation; I'm highly doubtful. There's more elegant ways of achieving a post-peak car situation that will be implementable sooner.
This type of system *could* work in areas built for this type of transport network...and that's about it. Older urban areas or areas that are more densely built have other solutions which will better serve the public, private investors and government.
TL;DR: There are other ways of doing it that are much better suited to the context than these two/three systems networked together.