Perhaps, but we are still a long way off from fully self driving cars actually giving rides to replace taxis at scale. It's not clear how Uber gets from here to there without continuing to lose many billions per year... at which point the whole self driving car thing will be largely a commodity that others could easily run themselves.

Right now Uber is really just proving that if you burn billions and billions you can create a "market" for your "product." It has yet to show that it's a real company with long term prospects. There's a difference between losing money because you haven't hit scale (when Facebook was losing money in the early years) and losing money because you're at scale and just undercutting everyone to show you can make "revenue" or by investing in "innovation" that everyone and their brother is doing and doing it better than Uber.

Self driving or driverless? I don't see any significant savings for self-driving cars, you still need to pay the person and you have the significant added expense of fleet maintenance. Not to mention liability concerns, unless they plan in passing that off onto the 'drivers'.

Driverless cars on the other had and still decades away, whether the tech is ready or not. Society is not yet at a place that will allow the use of full driverless vehicles on public roads. It doesnt matter how much statistically better a driverless car is than a human, every single accident will be magnified. Driverless cars could be our first foray into truly autonomous robotics within the general public, it is not something that will occur lightly.

For established tech and car companies, both Uber and Florida have basically no rules about testing and implementation of autonomous vehicles. Essentially the only rule is 'don't make us look bad, guys'.

Google is a league ahead of the competition with their Autonomous OS, and they're getting close to implementing robotaxis commercially in limited capacities. However, Chris Urmson, ex CTO of the self driving project doesn't think they'll be able to go everywhere a human can for ~30 years.

Uber's autonomous program is looking pretty sketchy at the moment, but now that they have Anthony Levandowski to play Darth Vader to Kalanick's Emperor Palpatine, the project is in capable hands.

Uber's robotaxis need only augment their human network of drivers through the mid-term, and as the capabilities of their Autonomous OS improves they take over more responsibility. Still, this is capital intensive and it'll be still be many years before autonomous vehicles have any measurable effect on Uber's bottom line, and even then, Uber will still be facing competition from other companies with robotaxi networks in development, including but not limited to GM/Lyft/Cruise, The Tesla Network, Waymo, Zoox, and any Automaker willing to license from the Mobileye/Delphi/Intel partnership.

Uber is currently playing hardball with the California DMV, in some sort of attempt at asserting their dominance over the law. I'm not sure if they've got some kind of legit strategy here or not, because I'm skeptical they'll have as much luck leveraging popular public support against regulators in the Autonomous Vehicle space with as they've had in the taxi industry.

It's also an inevitability that Uber's labour will organize for better bargaining power. Legislators seem to side with drivers as 'employees' rather than 'contractors' in the few regional cases that have made it through the courts.

Nonetheless, these incomplete financial reports that have been leaking out make it easy to spin a negative narrative about Uber's future prospects. That narrative may be on the mark, but we can't know for sure.

> The end game is a company with self driving cars, having SIGNIFICANTLY less overhead than they currently do with human drivers. And, probably a ton of technology patents to boot?

This is never going to happen. This is a pipe dream. Self-driving cars are so far away from being a reality that you might as well assume you'll never live to see them, and Uber will certainly expire long before they arrive on the scene.

In this particular iteration of the game, what exact benefit does running a $800M losing ride-hailing service confer?

If the entire value proposition is "taxi using self-driving cars", then surely they would be much better off dropping the money burning ride-hailing operation with cars they don't own to spend more on R&D...

> what exact benefit does running a $800M losing ride-hailing service confer

Market share. Or more precisely, establishing market dominance by undercutting the competition.

It's an app. If competitor X beats them to the self-driving-punch, I'll simply install their app and start saving. A first-to-market competitor will have 747s full of money landing on their doorstep begging them to take it.

There is simply no scenario where having a network of poorly paid drivers, their heterogeneous fleet of unmaintained cars and a load of debt is an advantage in a rollout of self-driving cars, the first useful version of which will surely be limited to one particular manufacturer, model and sensor suite.

> If competitor X beats them to the self-driving-punch, I'll simply install their app and start saving.

And Uber couldn't double down on their existing strategy of heavily subsidised rides to undercut self-driving cars? I mean, this is assuming that self-driving taxis are cheaper off the bat, which kind of ignores the fairly large capital expenditure of buying a fleet.

> A first-to-market competitor will have 747s full of money landing on their doorstep begging them to take it.

I don't think it's that simple. That's not taking into account how difficult it would be for Competitor X to build enough of a network to rival Uber, and how easy it would be for Uber to replicate/steal/buy the tech and steamroll Competitor X. Having superior tech and being first to market are not enough alone to trump market share.

>The end game is a company with self driving cars,

As it seems that is their only chance at profitability, then they should be plowing a lot more of their capital into driverless R&D.

I think they're already putting everything in there they can. They're probably the most aggressive developer of self driving capabilities (i.e. directly testing in the wild).