IMHO, autonomous cars actually breaks Uber's model. It changes it from taking a % off the top of a market to a capital and maintenance intensive business. Only seems to work if they end up with some sort of franchise model. Sort of like how Coke has a bunch of distributors.

That's about right. Running a fleet of autonomous vehicles is an expensive proposition, and probably will be for years after they start working. Waymo outsources some of that to Avis Rent-A-Car, which already has garages, maintenance facilities, parking lots, car washes, etc.

It's quite possible that when automomous vehicle operations start working, the big players will be car rental companies. It's a natural extension to their business.

I have a hard time believing that anyone on HN legitimately things that companies like Hertz and Avis will be able to become efficient in the ridesharing market before one of the two TNCs become efficient in the fleet management market (which is a subset of the rental car market that overlaps with TNCs once they choose to have their own fleet.

One set of businesses here has modern engineering practices and the other set are dinosaurs. Furthermore, rental companies have more employees dedicated to sales than to fleet management and their entire current customer base is rental car consumers. They literally need to redirect money to R&D long term where that R&D serves no benefit to their current customer base. Wall Street will punish the stock of rental car companies short term for spending R&D on becoming ridesharing businesses that serve a customer they don't yet even have mindshare with. With the TNCs however, wall street won't see any problem with them building out their own fleet. The only condition where wall street looks down on the idea of own fleet management is if they believe a franchise model with third party fleet managers make sense. Either way the TNCs own the relationship with the customer and therefore has all the power.

I would not be so sure. Avis, in particular, is actively pushing for a flow where you book online, use "secret" promo code to get full insurance for basically zero, pick up car directly on parking lot with keys inside, and drive away. It takes literally 5 minutes and a single employee to check your driving license and contract. Hertz is trying to do same, though less effectively. Their websites and apps are still horrible, and it's perplexing why, but the cost of building decent apps will be immaterial compared to the cost of the vehicles, and from that, building actual on-demand short-term rental network does not seem so unlikely.

I suspect this is what Waymo plans to do with Chrysler. Waymo will be the brand of AI driving the car, and probably own the app/marketplace, but the car companies will take care of building/capitalizing the fleet (and probably maintaining it, too).

The upside is that they keep 100% of fares, benefit from the economies of scale of having a large fleet, and can run the cars 24/7.

Cars can be leased, reducing capex, and maintenance can be handled by specialised companies.

Overall it seems that this upside far outweighs the downside.

Do they need to own the car. They could just run not in use consumer cars.

Are people prepared to do this?

I get my car to drive to me to work and then it goes out and operates like a taxi while I'm working. I'd be worried it comes back damaged or vomit all over the place.

No thanks.

I am not smart , but before the phoenix accident that statement may have some basis but after that the uber autonomous car is tainted to an extent couple that with theft of intellectual property from Alphabet autonomous trucking and the subsequent deal made, it is hard to argue Uber is going to lead the autonomous revolution when it happens. Their current moat is their driver network (one of the moats), once that is gone Waymo and Tesla can pretty much pull what Uber does.

Waymo/Alphabet/Tesla are one push notification to their apps away from building a driver network.

I've seen Waymo's cars driving around plenty of times in Arizona. The ones without a human safety driver are so rare I haven't seen one in person yet. If they were actually confident they were close to FSD they would have more cars without human safety drivers.

While I can understand why autonomous cars can improve the efficiency of the market, unless Uber is able to gain semi-exclusive access self driving cars alone I don’t understand how this will fix their economics.

Let’s pretend Waymo has 100% safe, easy producible, street legal self driving cars available today and everyone else was clearly several years behind. I see this would play out in a few ways:

1) Would Waymo actually want to spend billions to buy enough cars to maintain their own fleet in most cities in the world/US so that you could quickly and reliably hail a self driving car (and spend on marketing / referrals to grow marketshare). Would Waymos/Google investors actually allow them to lose this much money to start a new business that has historically been so unprofitable?

2) Would Waymo just sell cars to consumers/rental car companies, who would then deploy the vehicles to uber/lyft. In this scenario wouldn’t the cars be even more price sensitive about automatically switching between driving for different platforms and optimizing for the cars time. Wouldn’t consumer still open their uber/Lyft apps, check for the lowest price and then take the cheapest ride? Isn’t this fundamentally the same economics issue (unless the rides perhaps get so cheap that consumer no longer bother to check)

Any savings from not paying a driver would probably be taken up by maintaining and deploying their own vehicle fleet. And that's assuming the tech is ready in a reasonable timeframe and doesn't cost a ridiculous amount.

Well as someone who works in machine learning, I'm not convinced autonomous cars are anywhere close to becoming a reality, so they could have to survive for quite a long time.

2015 would like its hot take back.

I think the surplus from self-driving tech is going to go to the owners of self-driving IP and consumers.

Uber doesn't need to be profitable, they just need to survive until their investors can cash out

Who's going to give them more captial to build autonomous cars?

They'd need the following to really make it with autonomous cars:

1) Own the technology, not lease it

2) Have the capital to build out a fleet

3) Have the technology before others for long enough to build a moat.

Why do we believe that Uber would be the beneficiary of autonomous cars? They aren't developing them anymore, so if Waymo pushes them first why wouldn't the Google Maps app simply replace Uber, eg.