Autonomous Cars Break Uber

Why wouldn't Uber just switch from being a marketplace bringing together riders and drivers, to a marketplace bringing together riders and autonomous cars (and their owners)?

I suppose a well-financed rival could buy huge fleets of vehicles and compete directly with Uber, but Uber doesn't have to make that capital commitment. They just continue to be the middle-man between the riders and the cars, except the cars are no longer managed by their drivers, but by their owners (whoever they might be, individuals or large corporations).

Uber's big advantage is that they've already created the market. This is Paul Graham's idea about not automating in the early days of a startup taken to the extreme.

Well, it would break Uber if it weren't really clear that Uber is aware of this and wants to get in on it as soon as they can. Digital cameras broke Polaroid, but if Polaroid had gotten in on it early they could have been a dominant force in an emerging market. I don't think Uber is Polaroid here.

The key change for Uber when we have completely autonomous cars is that we will go from a two-sided marketplace matching drivers to riders to a much simpler retail model: riders buying rides. Whatever mindshare and network effects Uber has in the marketplace model will fall in importance, and the new king in the space will be the company that can deliver rides cheapest and fastest.

Uber is way behind Google and even Apple when it comes to calculating the best routes from place to place...

This is completely wrong. Self-driving cars don't break Uber - self-driving cars will help Uber reach its full potential. The assertion that Uber is a service that matches riders with drivers is completely wrong. What Uber does is match passengers with potential rides. As far as Uber is concerned, today it is an unfortunate reality of technology that the only way that ride is fulfilled is with a driver.

Drivers represent a huge cost in Uber's operation, and just a cursory glance at Uber news reveals that drivers have also been the greatest threat to their legal legitimacy. Uber is as highly valued as it is because it is one transportation business whose model is completely safe in the autonomous vehicle world. Uber will operate in exactly the same way it does today -- we just won't have anyone up front.

Ultimately the operating profit of a vehicle has to create a return on the capital outlay. Uber's current network effect guarantees a high utilization as autonomous vehicles are introduced.

higher utilization -> higher hourly operating profit -> margin to decrease prices -> unprofitable for new entrants -> barrier to entry

Without an existing network, the (the considerable! $Bs) capital need to buy a fleet of autonomous vehicles is prohibitive for even a determined new entrant.

(and so, as expected, the writer is wrong)

"[Uber CEO Travis Kalanick] recently told me that in 2020, if Telsas are autonomous, he'd want to buy all of them. He said all 500,000 of estimated 2020 production, I'd want them all, but he couldn't get a return call from Elon." — Steve Jurvetson

Kyle Samani completely ignores the diagram he spent so much time talking about. He's convincing himself that being able to park his $200k autonomous car a few miles away in free parking—summoning it to pick him up when he needs it—means he won't need to use a ride hailing service.

I wonder if he will still feel that way when he's been standing at the curb for half an hour waiting for his personal car to arrive, while the rest of the (non-car-owning) dinner party have hailed autonomous vehicles and departed within three minutes of leaving the restaurant?

Given the choice between buying a $200k car, or paying a few dollars every time I want to travel somewhere, I think I'd prefer the paid rides. The $200k could be invested in something productive in the meantime.

Uber is a dispatch and marketing system. The vehicle and driver are fungible parts of their machine.

Autonomous vehicles can enhance Uber's ability to offer travel with lower pricing than traditional vehicles and dispatch. Especially once they have goodwill built up.

Nothing will change instead of drivers buying a car and joining uber. People will buy autonomous cars and register with Uber.

I'm surprised there's no mention of Uber's Advanced Technologies Center. IIRC this is exactly why they made it.