[wkcamp]

Just to add on, think of how much more centralized the company's model becomes. A self-driving car probably won't depend on third-party drivers and Uber can start intaking a lot more cash from its services (of course, maintenance and such costs may increase).

[wpietri]

You're also neglecting capital costs. I think Uber's major business innovation is shifting almost all of the cost and risk onto drivers by selling them "independence" and "great money". Previously, cab companies had to buy and maintain fleets of expensive cars; the biggest risk a driver took was the cost of one shift's rental. Now some are in hock for tens of thousands of dollars. And possibly millions given that they are no longer covered by taxi company insurance.

Honestly, I think it's more risk than opportunity for them. Once driverless cars are ubiquitous, the barrier to entry will be pretty low. Currently to compete with Uber you need to sign up a bunch of drivers, trick them into paying most of your expenses, and then pay them just barely enough to keep them from quitting. But if all it takes is money for cars and the ability to buy an app, anybody with a strong brand and access to capital can go after them.

Imagine, for example, you're BMW's CEO. You can build custom cars tuned for taxi service. You know the in-car software very, very well. You have a mess of industrial designers and deep research on customer motivations. Whenever normal-market orders dip, you just keep building for your taxi fleet and occasionally declare war on Uber in a new major city when you can flood it with shiny cars. You already have a service network and get parts at cost. You have incredible access to capital, great relationships with millions of existing consumers, and one of the world's strongest brands. Wouldn't it be tempting to say, "Fuck it, we've been in the transportation business for 100 years! This is just the next step."

[tptacek]

This is not really how a lot of cab companies work. Drivers can end up leasing cars or medallions from cab companies too. Cab drivers get screwed by cab companies, which are, contrary to the narrative that needs a counterbalance to Uber as the "bad guy", not corporate model citizens.

[wpietri]

I'd be interested to see some data on this, but every single cab driver I've talked to in San Francisco was on the per-shift model. And when I lived in Chicago, same deal there. That'd be dozens in each place.

I also agree that cab companies are kinda terrible, but that doesn't make Uber any less terrible. If anything, I think it's a sign to us that we should be especially careful with Uber: if a few local companies competing can end up all being pretty bad, one nationwide monopoly run by ugly people can only be worse.

[tptacek]

The cab drivers in SF always complained to me about their relationship to the dispatcher, but I'm not sure I ever grokked that.

Chicago cabs, I get an earful on leases and minimum number of hours they have to work to pay for their cab/medallion.

You can be on a shift AND in hock to the cab company; the two aren't mutually exclusive.

To me it seems like at least with Uber you can set your hours, and there's transparency regarding your capital costs: you buy/lease the car yourself, so no medallion or other lease shenanigans.

I'm not in love with Uber, by the way: I think Uber should have to comply with a lot of taxi regulation that they don't currently comply with.

[wpietri]

It has been 15 years since I've lived in Chicago, so it must have changed. At the time from the cab companies I heard about per-shift leases and drivers were very focused on making the nut, that being the amount to pay for the car's use that (12-hour) shift. But I could see why cab companies would force a shift to longer commitments, as the drivers were pretty strategic about only picking up shifts where they were sure to make money. Great for drivers, but that could leave a lot of cabs sitting idle. And now that you mention it, I recall a few drivers who owned their own cab/medallion, and would time-share it with a couple of driver buddies; they were much more focused on maximizing road time. So I think I was too hasty above.

In SF I've certainly never heard a driver complain about a contract with the cab company or minimum hours, and given the number of complaints they have when you get them rolling, I figure I would have heard. But I'll ask next I get the chance. And the complaints about dispatchers have gone down dramatically now that cabbies have other ways of getting passengers. Flywheel in particular is a big winner for everybody, and I wish some taxi commission had been forward-thinking enough to force a whole city into a universal dispatch with a clean API.

[pvg]

In SF every driver, except for the tiny minority who own a medallion pays the gate fee. Since the medallion licenses a cab to operate 24/7, medallion holders also lease out their medallions if and when they don't use them personally.

The MTA sets the meter rate, the gate fee and both the number and sale/resale price of medallions. It seems it almost doesn't matter how evil or virtuous cab companies are since the system doesn't appear to have much flexibility, room for competition or self-correction.

[minthd]

The BMW scenario is interesting ,but:

1. Will ride sharing still be a thing with self-driving cars ? Because it could have a strong network effects.

2. Will consumers prefer the most tested vehicle, let's say - Google's ?

3. Are BMW fit to sell a commodity product more than others ?

[wpietri]

My point is not that BMW will win. My point is that automated cars make it a low barrier to entry. Many, many companies could reasonably think to seize a piece of the transportation-as-a-service market.

The list starts with every car company in the world, because a) their execs all went to the same management schools and learned the same lessons, and b) they all have the necessary factors for market entry, which I'd say include 1) access to capital, 2) access to cars, 3) ability to make software, and 4) a strong consumer brand.

But the list definitely will not end with Uber plus the car companies. Transportation is 8.5% of our total economy, and automated vehicles will totally change it. Hopefully they will also reduce it to 4-5% of our economy, so there will be a lot of bloodshed, but in the chaos there will be some great opportunities.