> they already had a built-in customer base using them for ordering point-to-point rides I never understood the value of that. The value is in the driver netw

> they already had a built-in customer base using them for ordering point-to-point rides I never understood the value of that. The value is in the driver network, which is hard to build with human drivers, but trivial with self-driving cars. Given a driver network, building a customer network is trivial. You can have a large fleet that moves to a city that you want to bring up. The novelty of a self-driving car will help find users, the sudden large presence of highly recognizable, branded vehicles will automatically serve as a billboard for your service, an offer of free rides does the rest. Once people realize that the self-driving option is a) cheaper b) faster (because of the excess of cars) c) safer (because there is no unpredictable human driver that could be a bad actor), human-based competitors will have a very hard time. Once the city is brought up, they can leave the number of cars that they deem appropriate (gradually worsening service), and have the rest of the fleet drive to the next city. What does Uber have? An app (easy), a way to handle fraud and abuse (totally changes and gets a lot easier when you remove the driver), network effect (see above). Google has a special advantage, the maps app. If they manage to not ship the org chart, they can add a "order Waymo" button to navigation. Boom, everybody has their ride share app already installed on their phone, and many of the users have payment data in their account.