The issue is whether any of that will matter if self-driving cars come into the picture sooner rather than later. Uber's key advantage is that there's a huge network effect that's hard to overcome with humans. Drivers want to use the app that riders are using. Riders want to use the app that drivers are using.
If a company could go out and finance a few hundred self-driving vehicles, it would be easy to build up market share for competing services with lower margins. Where I am, Uber has far better coverage than competing services. But if it were easy for alternatives to cover equally and with lower margins, riders would easily check the cheaper services.
While Uber is certainly employing robotics engineers, so are the auto manufacturers, Google, and possibly others. I think people use Uber for the same reason that people use Facebook - it's hard to get all the other humans to use other services when the value of those other services is how many people are using it. But when self-driving cars allow anyone with enough money to finance enough vehicles that they're offering comparable wait times, Uber's network effect disappears. Sure, it will still have brand recognition and some amount of loyalty. But it won't have the kind of barrier to entry that it has today.
Uber is winning, but it's also subject to a huge disruption by self-driving cars. If one can purchase and operate a self-driving car for $500/mo., you only need to do $17 worth of rides per day for that to pay you back. If Uber tries to keep its margins up, there's no reason someone isn't going to come into the market with some VC and drive those margins down quite a bit. Lyft, Sidecar, Split, and others all face a huge challenge of getting drivers to work for their service before they have riders and getting riders for their service before they have drivers. If they could just grab some VC to pay the finance charges on the vehicles until ridership picks up, that's a game changer.
Uber goes from Facebook with a wonderful network-effect moat to Amazon who has to make sure that margins stay razor thin to maintain a quarter of the market. I'm not saying that Amazon can't do good things, just that Amazon doesn't get high margins and if it tried to up its margins 10%, many people would shop elsewhere. With self-driving cars, someone is going to have near-zero margins and if Uber is 20% above that, people will switch. In fact, if Uber had margins of 60-70%, why wouldn't Amazon (who has a reasonable amount of robotics expertise with Kiva and their drone research) step into the market at significantly cheaper prices?
Heck, Amazon could have an automated delivery fleet of cars with drones handling the car-to-door problem and then Amazon could use their "excess capacity" much like they did with EC2 to drive people Uber-style. During the 10-4 day (when there's less demand for vehicles), they could deliver Amazon packages. From 7:30-9:30am and 5-7pm they could do rush-hour and 7pm-4am do restaurant and bar patrons. That's probably a lot more use out of those vehicles than one could get simply from people wanting an Uber and that would lead to lower prices.
Sitting here in the cheap seats, I guess I wonder how Uber would fare against that business plan. Amazon accepts thin margins and has a giant use case for their vehicles for the hours that self-driving Ubers sit idle. I'm not saying that Uber isn't filled with smart people. The issue is simply that Uber's current success is probably mostly due to the network effect. They might still have future success, but it's going to be tempered by the fact that competition will be a lot easier.