Driverless cars don't exist and there's literally zero evidence that Uber has a stronger claim on that market than any other aspiring company.

Engineering for an app like this is non-trivial but hardly insurmountable.

The one advantage they do have is network effect, the more drivers and users they have the quicker it is to get a ride and more lucrative for drivers. That does matter, the question is how much.

Driverless cars at least half a decade if not a decade away from widespread adoption in cities. Who's to say they'll last that long? What does "state of the art engineering infrastructure" give users? I don't recall Lyft's app ever failing on me.

If anything Driveless cars is a huge threat to uber. Buying the cars would be a massive capital investment. Right now they rely on their employees/contractors to bring a 20k+ asset to work with them.

Someone like Hertz or a General Motors already have fleets and know how to manage them. All they have to do is engineer an app.

What is easier? A nationwide vehicle fleet or an app that connects drivers to riders.

Driverless cars will happen at the same time for everyone. Uber's competitors will get them at the same time they do. It will not amount to a long-term or even short-term advantage.

Finally, Uber get's pushed out of relevance by driverless cars, from BMW, Mercedes-Benz, Chevy, Ford, etc., who will ultimately never let them have this market.