> That cabs only really exist in densely populated urban areas should tell us something about the economics of this business model.

yes, it tells us that human cab drivers are prohibitively expensive. robot drivers don't need to paid by the hour though. the real disruption to your car isn't a ride-hailing app on your phone. it's robot driven cars. ride hailing apps are necessary piece of predecessor infrastructure for that transportation model. it's the beginning of it, not the end.

In my opinion self driving cars are a shiny object to distract VC's from Uber's massive operating losses. When they start making money with self-driving cars on a large scale then I'll take notice.

Uber won't even be necessary. If it becomes cheaper to "rent" the service of the car (i.e. Uber/Lyft with autonomous vehicles), then car manufacturers can just cut out the middle-man. Uber/Lyft would serve no purpose anymore unless they somehow make it cheaper (which would be practically impossible).

The landscape of the future is car manufacturers offering their own "Uber" like service and that service replacing the traditional ownership model. That doesn't mean Uber doesn't have a place before then though. Like good little capitalists, the VC's will cash out and make a ton of money while the real innovation and lasting impact happens elsewhere. So, yeah, Uber should blow up when self-driving cars become a thing but they'll eventually fade away. Unfortunately, that won't be for a long time.

EDIT: Oh, and yeah, people are also going to still be buying a shitload still anyway.

wait 5-10 years? there aren't any self-driving cars in commercial use right now. VC investors often have long term timelines (depends on the fund and it's objectives). It's not exactly a distraction tactic to say "our timeline projects out for a decade and we expect to be operating in the red until the technology and politics aligns and comes to fruition, at which point we make infinite money forever."

Yes but most funds are marketed as repaying in 10 years and, even if some take longer, the VC presumably does not want that (since it, presumably, prejudices their chances of raising their next fund).

Expect an even bigger fund raise where they hope some other idiots, with even deeper pockets, will keep picking up the tab.

Can Uber make it 5-10 years losing $3 billion per year?

Removing the driver reduces the price significantly, so the business model of cars on demand is probably viable outside of dense areas with self driving cars.

That being said, for Uber to win the self driving car race they have to beat Google, Apple, Tesla, Didi and the automotive incumbents as Sarah Lacy has been pointing out for at least 12 months.

they don't have to beat those companies, they have to negotiate good rates for vendor services with those companies.

What does Uber have that would not be feasible for any of those companies to build?

Uber doesn't lock in customers or create any switching pain whatsoever so their current customer base doesn't given them much of an advantage. Their software is slick and robust but it's not so technically advanced that a competitor couldn't create their own ride-calling and dispatch system (as evidenced by Lyft).

Have you seen how bad and outdated the software is in even high end, expensive cars?

I wouldn't hold my breath that any auto maker besides Tesla could compete with Uber.

> What does Uber have that would not be feasible for any of those companies to build?

usage data. they know where people that use their app actually go on a daily basis, and how much they're willing to pay for that service.

There other ways to get traveler data, people have been modeling this for decades.

Or be acquired by one of them.