The second point is that there are two keys to making money out of this: Have self-driving cars, find customers to rent the self-driving cars. Uber is already established in the second part. So it might be true that another company buys the cars and tries to run the business, but uber has the established customer base.
In my opinion the future of self-driving cars inevitably ends up with one of the big players buying uber to allow them to monetize their fleet.
Also, I never understood what kind of problem companies are trying to solve with self driving cars to be honest. Traffic in cities will be the same, if not worse. And I consider driving as one of the things I really enjoy doing. Why take that away?
So it has to buy parking.
What happens when they break down?
Servicing.
How often do they have to replace their fleet?
$3 billion every 5 years.
> So it has to buy parking.
Rent parking spaces in people's driveways! :-)
Let's assume, for the sake of argument, that the concept is viable in urban areas within 5 years and Uber puts a massive fleet of self driving vehicles on the road.
While it's certainly true that they wouldn't have to pay drivers it would be a massive shift in capital investment.
The "genius" of Uber is shifting the necessary capital investment for hundreds of thousands of vehicles onto the drivers.
Assuming one car costs 30K (which I think is on the very low side) it costs 3Billion to put 100'000 cars (which also seems very low with Uber's global ambition) on the road.
That's an upfront investment currently incurred by the drivers.