Hacker News Reader

Comment by SilasX | Hacker News Reader

From the summary in the child twitter link:

>We hope in the future there will be driverless cars and that we can then make money because no drivers but other people are developing them too.

Nice! It's common to think that self-driving cars will be Uber's salvation, but even then, they'll have to come with other SDC providers, so there's no moat. They'd have lower costs, but so would everyone else, and so they'd have to cut fares proportionally.

In theory, they can have such a lead time on a workable SDC that they get a great moat until everyone catches up. But that would be ~2 years at the absurdly optimistic end; not enough to pay back the costs of the program and throw off the huge returns investors demand. And in reality, Uber is lagging on this.

View on HN

The self driving car idea won't help Uber or Lyft for two reasons. First, neither company has the tech experience to build one. Second, and more importantly, both companies would have to buy and maintain the cars which isn't financially viable.

Indeed, part of the ability for the driving "gig economy" is the masking of risk and liability from the corporation onto the individual driving humans for car expenses and depreciation. The myriad of humans thinking Uber/Lyft are job have yet to get caught up the the expense of the car itself.

Uber/Lyft both hope they can get to self driving cars before the drivers figure out the true cost after the car they provided to Uber for free is involved.

And in the (imho unlikely) scenario that fully autonomous taxis are a viable mass market opportunity in the foreseeable future, both companies would find that rivals who actually do make the cars and have networks of maintenance centres are in a much better competitive position. Making the pricing and matching app isn't the difficult part, and Uber's existing driver base ceases to have value

They're hoping you'll buy a self-driving car and then let them pay you a minimal fee to use it when you're not.

The problem is that self-driving cars will change car ownership needs, and that there could be entirely different companies in the space when it does finally come around.

I'm not sure I buy that scenario...

Is there any other market where random consumers are basically acting as financiers for a commercial fleet of equipment? What's the reason for the fleet operator to involve those customer-owners at all?

In terms of immediate cost, it will be cheaper as a customer just to use the ride sharing service yourself, since you won't have to pay car payments you'll just pay the rideshare fee. Especially if the self driving software gets priced in the $10,000+ range as Tesla has suggested it will end up.

So anyone in the "budget" category will just be buying rides not vehicles. They'll have to wait 5 minutes for a vehicle, and they won't get to have their vehicle of choice, but they also won't have to deal with the hassle of running a mini carshare business, whatever that will entail.

Most people in the "premium" category isn't going to want random drunk people fingerblasting each other in the back seat of their nice car, so that cuts out the top of the market.

That leaves what... people who want to save some money but also really want to own their own car, but also are comfortable sharing it with random people?

Seems like a weird segment. Why wouldn't the fleet operator just buy cars and own them? They're going to be burning through these vehicles every 5 years anyway, what's the point of parking them in a random owner-operator's garage and having them taking it out of the fleet when they feel like it?

Indeed. And surprisingly, no one[#] has brought up that there is no such thing as as a national (or even regional) taxi service. And why is that? It's because the economics of fleet ownership don't scale to that size. If they did, we'd have a single nationwide taxi service already. In 100 years of taxis, no one has done that.

[#] Except Hubert Horan who destroyed all their BS with his series, "Can Uber Ever Deliver?"

Yeah, but arguably smartphone technology reduced admin costs and raised the size of the economically optimal taxi provider. That has been around for less than a hundred years.

Edit: Dangit, I thought my reply got eaten, but I was looking at a different branch. See the rephrasing here: https://news.ycombinator.com/item?id=19880565

Smartphones didn't reduce the cost of dispatch. Look what Uber spends on app development! The big costs in taxis are maintaining the fleet, a cost which Uber externalizes to its drivers, some of whom now have to sleep in their cars.

They most certainly did. As expensive as that software is, it’s very little when amortized over the rides. It eliminates the overhead of finding a location and transmitting the info, and delivers higher value in that it gives real time data about where the car is, and lets you pay more conveniently.

Also: while the drivers have physically taken over maintenance, Uber can’t push the economic cost on to them, because they have to pay enough that drivers want to continue to work even given those costs. If anything, maintenance is cheaper when handled by a big org.

To the extent they realize a savings, it’s because they’re using the spare hours of existing vehicles and don’t have to take on all the dead ours they’d have from owning the cars not because they’re getting maintenance free in any meaningful sense.

The reason there isn't a national chain is because municipal laws fragment the market. You can typically schedule a cab for pickup in rural America (I've done it before), it is just expensive because you're also paying for a person's time to drive out there and back.

That's my point - autonomous car fleets will change car ownership habits. Increased inexpensive, on-demand transportation means less people buying cars, and the people who do end up buying won't want to share with strangers.

I doubt that people who have to sleep in their cars will be able to afford an autonomous vehicle.

It's not unrealistic to roll that out as a profitable service.

It is unrealistic to expect that to be a moat-profitable income source for Uber, when you can rent that same SDC to Lyft -- or Tesla, for that matter [1] -- who will offer competitive terms.

[1] If they offer it for their own cars, they could probably extend the platform to allow outside cars, perhaps with different branding.

Yep, the landscape will probably be entirely different and we'll be talking about the next ridesharing unicorn that managed to corner the self-driving rideshare market.

> neither company has the tech experience to build one

They have the experience to distribute software to millions of devices on a regular basis and run a logistics platform and a consumer app.

That's something automakers will struggle to do.

Automakers can build cars.

That's something Über will struggle to do.

Seems like a fairly straightforward partnership to me. There are many automakers and most of them (anything vaguely in the budget category) will be struggling mightily if robotaxis ever work.

Mightily struggling automakers like to make deals where they get to make autos.

The tech to distribute an app? You mean Google Play, lol? A teenager can get an app in there.

Uber can't afford to run a national taxi service at scale for the same reasons no one else has ever done so. The economics of a taxi fleet don't scale to that size. Owning their own autonomous vehicles would only further erode their non-existent business model.

No, I mean built the app, maintain it, build a relationship with users, get people to download it and tell their friends, all of that.

> The economics of a [self driving] taxi fleet don't scale to that size.

I'd be interested in hearing your analysis! I haven't heard that one. Tesla claims these vehicles will pay themselves off within two to three years once they can operate fully autonomously.

Hubert Horan has done an amazing teardown with his series, "Can Uber Ever Deliver?" But you don't need to be a transportation logistics expert (as he is) to see that there are no national or even regional cab companies. The reason is that, after a certain size, the business simply doesn't scale.

Arguably, smartphones, apps, and usage of the drivers' own cars have together reduced the administration costs, and thus raised the scale at which a provider can profitably operate. That interplay of factors hasn't been around for 100 years.

Although I've often speculated if you could have had an Uber-like service in the 80s, where you register with your credit card on file, and then call a number with a location to get an estimate pickup, and then they page a driver to get you, perhaps with an authentication code.

And maintain includes very regular things like cleaning etc. I can't imagine this is things they want to do so it will probably stay a network Tesla style, but with lower (hopefully) costs.

Reply on news.ycombinator.com