In all seriousness, the technology that we are very close to being able to realistically automate is rail in less than fully separated environments. The nature of operating on rails eliminates the sort of unpredictable ad-hoc problem solving that is going to be a problem for truly autonomous vehicles, and while the application is fairly specialized it is significant enough that there could be real money on it. Realistically the two use cases boil down to being able to get all the sort of operational cost benefits that go with light metro like systems (think Vancouver, or what's being built in Honolulu) that are automated and as such very frequent but have the kind of capacity more associated with light rail without the cost of full grade separation, and being able to automate more typical transit routes with less than full grade seperated routes (i.e. streetcar systems like in Seattle or Portland suddenly have a big advantage over buses beyond capacity and aesthetics - they can be driverless, a change that eliminates upwards of 60% of operating cost). For that matter trolleybuses might even be close enough to fixed guideway to solve a lot of the sort of problems that full automated cars would encounter, though this is more complicated and possibly introduces some more liability (there are legal advantages to being a train rather than a motor vehicle in most jurisdictions).