Stage 2, the transition to the self driving vehicle, can be won by whoever is able to supply self driving vehicles at scale.
It is plausible that Google and Apple could create a ride marketplace for vehicle manufacturers. You press a button on your phone and a car shows up. This would provide a survival path for car manufacturers.
Another plausible scenario is cities converting to self driving "platforms" exclusively run by single companies such as Uber or Google. In the shorter term this could be a winning play because it makes the self driving automation part a little easier. Longer term this doesn't scale so well. However, if a larger company is able to get big cities to sell transportation rights there could be long term lock in. There will be big opportunities here in the next 5 years as municipalities continue to become more strapped for cash.
Google, Apple, Uber, Tesla, and BMW are the companies I'm watching closest right now. The big question may be what was more important -- have the rider/driver network first, actually manufacturing self driving electric vehicles (electric is important because emissions and air quality will win cities), or having the platform users use to hail vehicles (Apple and Google.) Uber could have a big IPO and buy a car manufacturer outright but I don't know if that makes much sense. Buying up assets of a bankrupt or bleeding car manufacturer seems more opportune.
Sure, but if every company (think Tesla/Google/Apple/Amazon etc) eventually can build/buy autonomous cars and offer the same service, what competitive advantage would Uber still have? Their current network effect of matching drivers with customers is where most of their value is imho. If that disappears, then I really don't know why we still need Uber.
If I speak into my wearable iEarplug in 2020 and say "Hey Siri, get me a car", would I really care about it being an Uber car?