Hacker News Reader

I know that's the sales pitch, but does the economics actually work out that way? You mention subsidized ride hailing, for example.

> (personal cars are utilized <5% of the time, while self-driving cars can see >60% utilization)

How much of that 5% is commuting, though? If there are two one-hour long windows in the day where a lot of people want to make the same trip at the same time, the fact that cars are idle in the middle of the night or day doesn't help with that. And that's also going to be peak surge pricing time.

The time economics gets worse in non-suburban areas. In high density urban areas, it's already too congested to not take public transport. In very low density areas, you might hail a ride, but you've got to wait for it to become available and arrive.

> Self-driving is proven technology, see Waymo.

Only in certain locations, and still dependent on occasional remote operator intervention. Tesla have been promising for years and not delivered, and every year they don't deliver makes it less likely that they ever will.

I think there's room in the market for such substitution, but it underestimates how much people love their cars as a form of personal space and personal brand extension.

Your analysis neglects the effect of self-driving vehicles on public transit. Self-driving busses and self-driving vans will vastly expand available public transit routes. Many public transit routes are not profitable (because they require a driver who costs ~$20-30/hour). Self-driving busses and vans will enable these routes, and shift people away from private vehicle long-distance commuting.

This will have a number of benefits, including increasing frequency of public transit, reduced traffic, reduced long-distance transit costs, etc.

Waymo is actually viable in pretty much the entirety of the US, they are able to expand whenever they want, but choose not to, because they're too risk averse.

When was the last time you sat in a self-driving Tesla? Today it's actually really good. It's gotten so much better over the past 5 years. I can see Tesla's self-driving business being viable by 2030.

> underestimates how much people love their cars as a form of personal space and personal brand extension.

This is the most vocal demographic of vehicle owners, but in reality they are not a significant percentage of the population. IMO most people don't like driving, and would rather not drive.

When was the last time you sat in a self-driving Tesla? Today it's actually really good. It's gotten so much better over the past 5 years.

Yeah, but people have been saying this exact thing for literally a decade. "Have you tried v18.419.18304, bro? It's so good, another six months and FSD will be solved!"

Lunacy.

It's very relevant. In my experience, I've noticed the majority of FSD critics (either for Teslsa FSD or Waymo) have never actually tried riding in one, or last tried it nearly a decade ago when the technology was immature.

Products change. Apple maps was a terrible product a decade ago, but today it's arguably better than Google maps.

pjc50 never replied to the question, so I'm going to take that as a 'never tried it or haven't tried it recently'.

Yeah, maybe it's different now, but the Tesla stans burned their credibility many years ago.

Also, the evidence still suggests that FSD is nowhere near as safe and capable as Waymo, so I find it pretty uninteresting tbh. Tesla seems like just a washed-up former side project of a drug-addled billionaire who has moved on from trying to build things to trying to destroy them. You can spend your money however you want, but I'd rather drive myself than give Musk a penny.

If this self driving future is better, people will be driving more, not less. Add dead heading and milage will increase even further.

If cars still get about 200 000 miles of life like they do now then we'll have fewer cars, replaced more often, so still requiring the production of a similar number of cars.

We'll need million mile cars to reduce that, and those don't appear to be coming from Tesla.

> If this self driving future is better, people will be driving more, not less

People won't be driving at all.

> still requiring the production of a similar number of cars.

Doesn't address my point, that consumers won't be the ones buying cars. These cars will be sold to self-driving vehicle fleet managers, not consumers.

Obviously, cars will still be made, but not consumer cars. The consumer car market is dead.

> those don't appear to be coming from Tesla

Tesla just killed its consumer vehicles Model X and Model S. Tesla shifted focus into the Cybercab & Cybervan, both self-driving only vehicles, which don't have steering wheels.

> > If this self driving future is better, people will be driving more, not less

> People won't be driving at all.

Poor choice of words. There will be more cars on the roads, not fewer. Somebody will have to build those cars.

> Obviously, cars will still be made, but not consumer cars. The consumer car market is dead.

But the car market is larger, not smaller, and that's what matters. Fleet vehicles are generall not cheaper than consumer vehicles; a new city bus is now $1M and not because city transport authorities are swimming in cash.

> Tesla shifted focus into the Cybercab & Cybervan, both self-driving only vehicles, which don't have steering wheels.

Which is only one portion of the picture. Have they spent any effort on million mile batteries? Have they spent any effort on million mile suspensions? Have they spent any effort on million mile interiors -- nobody wants to get into a taxi with > 200 000 miles on the odo, but million mile trains and buses are fine because they're made of hard wearing material that is easy to deep clean.

Taxi drivers generally run Toyota's and Mercedes' for a reason.

Do you live in a dorm with a shared bathroom? Most of your home is only used a few times a day.

People buy cars because it's a little bubble of home away from home. They store their stuff in there, it smells like them, and they don't get stranger's vomit on a seat when they want to drive somewhere.

The "people won't buy cars because self-driving" take just completely ignores the human nature.

I understand your argument but you are neglecting the numbers.

The cost per mile for a personal EV is $0.75/mi.

The cost per mile for a shared AV EV (at scale) is $0.40/mi.

The cost savings are extremely significant. I would agree with you if these numbers were close (maybe 10% cheaper just to rideshare everywhere) but it will be nearly 50% cheaper to rideshare self-driving than own. Tens of millions of people are going to switch.

Not to mention -- as more people switch to shared AV, the cost of ownership for humans goes UP even more. Fewer human drivers mean higher insurance rates, fewer personal cars are sold so the cost per car increases, personal car repair shops close and the cost of individual car repairs increases. Etc.

You can make even larger savings by moving into a dorm tomorrow. Are you neglecting the numbers?

Are you neglecting the numbers? Dorms barely exist in the US. You couldn't move into a dorm even if you wanted to. They were effectively illegal to construct since the 1930s, regulated to death by the 1950s, and only recently rules on SROs were relaxed to allow them.

I've seen "hacker houses" try and fail to make dorms. One I went to put three bunk beds in the basement, with 12 people living in the house. They eventually got shut down by the city for violating occupancy laws.

If dorms were legal, more people would use them.

Sure, more people would use them, but it doesn't mean that the majority of people would trade normal housing with dedicated bathrooms (that are used only for a fraction of the day) for that. Normal housing market didn't die when dorms were legal to construct.

The savings that can be made from living in a tiny box with a shared bathroom instead of a regular American house are extreme and are much bigger than savings from not owning a car. Do you seriously think that will convince people and we will see the death of dedicated housing units in the near future?

I ran the numbers. Car ownership today @ 15,000 miles per year costs $1,003/mo. Traveling the same distance with AV would cost $520/mo. Monthly savings are $483/mo or around $5,800/yr.

See I'd agree with your argument if the savings were small, basically nobody is giving up their car for 10% savings, but these savings are very significant. Car ownership costs are the second biggest expense for most households after rent, and self-driving cars cut this in half.

I can see many households downsizing, and instead of owning 2-3 cars, only owning 1 and replacing those trips with self-driving rideshare fleet vehicles.

Especially for the younger generation - teens today drive far fewer miles than ever, which skews the costs away from car ownership even futher. I predict many will not get their own individual car, they'll just rideshare.

> Car ownership costs are the second biggest expense for most households after rent, and self-driving cars cut this in half.

Living in a hostel with a shared bathroom can cut the biggest expense (rent) by more than half.

Would you?

> Living in a hostel with a shared bathroom can cut the biggest expense (rent) by more than half.

I don't believe you. Run the numbers for me for a private room at a hostel in NYC or SF, and get back to me.

> Tesla shifted focus into the Cybercab & Cybervan

Building on the proven sales success of the Cybertruck?

This doesn't change anything. Tesla is not leading in anything anymore.

I was in China a few weeks ago, and in some cities you can already get the equivalent of Waymo. There are also dozens of huge companies working on self driving there, with very friendly laws that make it easier to get training data and test things.

There are hundreds of companies working on robots as well, and many of them are already ahead when it comes to productionizing them.

Tesla entered a new market around a decade ago, back when they had little to no competition. For years, they were ahead of everyone. But now, everything they do has competition, and in most features/products that competition is ahead of them.

Their valuation doesn't make any sense.

The consumer car market will collapse 50-80% by 2040

Absolutely not. You're correct about Waymo having solved self-driving cars, but approximately no one is going to choose using Waymo over owning a self-driving car, just like approximately no one did that back in 2015 with Uber. Sure, a few bloggers in major urban areas did, but they represent a tiny, tiny fraction of the global market. People like owning cars, for a variety of reasons, and I don't think that's going to change. Even if I took Waymo 80% of the time, I'd still want a car. And I hate owning cars! It just beats the alternatives.

Many, many people will switch, because the cost difference is absurd. The cost per mile for a personal EV is $0.75/mi. The cost per mile for a shared AV EV (at scale) is $0.40/mi. People will save hundreds of dollars per months by not owning a car.

Millions of Americans can barely make their car payments, car payments + insurance are the second highest expense for most households, and I can absolutely see a huge percentage of this group switching to self-driving rideshare only.

I'd generally agree with you that people want to own cars, but for many infrequent drivers it does not make sense to own.

>Many, many people will switch, because the cost difference is absurd. The cost per mile for a personal EV is $0.75/mi. The cost per mile for a shared AV EV (at scale) is $0.40/mi. People will save hundreds of dollars per months by not owning a car.

What's the cost per mile at scale for a train?

The NYC subway has an average operting cost of $1.09 per passenger mile.

Also, trains are not directly comparable because self-driving vehicles are point-to-point, while trains are hub-to-hub.

Infrequent drivers make up a minority of Americans. I wish more people used public transit and cycling for transportation, but it's not a cultural practice outside of NYC and a handful of neighborhoods elsewhere.

> No, it's because Elon understands the model of selling cars for humans to drive will be dead by the end of the next decade.

Yes, that's Elon's claim.

I own a Tesla with FSD (hardware V3, so despite Elon's promises it will never work). I would never hire it out as a taxi. The risk of malicious damage is too great. I would have to insure for it, wear not having the car available when it happens, and the emotional impact of people destroying "my stuff" for the thrill of it.

The future of taxi's is a more traditional model like Waymo.