Hacker News Reader

Comment by pjdemers | Hacker News Reader

Tesla, and even more Uber, were priced as if self driving cars are almost here. The idea that services like Uber without the drivers would make car ownership obsolete drove those valuations. Now that self driving cars are many decades away, instead of less than one decade away, is bring stock prices down. I always questions how fast we would see self driving cars, not because of the technology, but because of consumer demand. There was a survey a few years ago of how much more people would be willing to pay for cars that could drive themself, and the median answer was $0. This is because quite a few people gave a negative number, that is, they would pay extra for a car that could not drive itself.

View on HN

> There was a survey a few years ago of how much more people would be willing to pay for cars that could drive themself, and the median answer was $0. This is because quite a few people gave a negative number, that is, they would pay extra for a car that could not drive itself.

I think there's an inherent distrust (which I share) of giving technology the power to end our lives with one bug. Trust in technology, in general, is pretty low right now, so that reflects back on perceived future technology where one bad steer or missed brake ruins your life. Although I happen to share this pessimism with the human aspect of driving as well, I've almost died before because I wasn't focused while driving (IRL nerd sniping basically), and I know it could happen again in the future no matter how much I want to be a safe driver.

The truth is, if full self driving cars actually happened, with safety statistics better than humans and with a low barrier to entry, you can bet that people would change their opinion real fast. Much like Americans who have a revelation after experiencing European or Asian public transit, the realization that they could have an extra hour per day that was previously spent staring at a road, would be a big motivator to change opinions on technology. Add in that three of your friends already use it daily and they're not dead yet, and you rode with them and it seemed like a safe driver.

>>with safety statistics better than humans

The problem is that will not be the expectation. It will need to be perfect or near perfect.

full self driving cars can drive a million of miles with no accident, but if on 1,000,001 miles there is a accident with a death it is all over the news, and people proclaiming the technology is terrible, unsafe and not ready. In that same 1,000,001 miles humans have caused far more damange and death but that is just "normal" so....

from a distance that's how people feel now. because self-driving isn't something anyone has access to, it's something that a few tesla buyers and a few tech companies (waymo, etc) have access to. So, from an average person's perspective, of course I don't want those rich/greedy assholes killing kids. The algorithms need to be perfect before they're allowed on the road

If every single car were upgraded to have full self driving tomorrow, with algorithms to perform lets just say 1% better than humans perform, I guarantee that acceptance would build overnight.

Well obviously. If a self driving car caused a fatal accident after just 1,000,001 miles it would be like 80x more dangerous than the average American human driver.

The reason we see summary dismissals of current autonomous vehicle technology is because even the best systems by Waymo and Cruise are literally orders of magnitude away from human-level. No discussion or societal cost-benefit is necessary when they are years away from reaching the bare minimum level worthy of analysis.

> The truth is, if full self driving cars actually happened, with safety statistics better than humans and with a low barrier to entry, you can bet that people would change their opinion real fast.

I think this is the wrong counter-argument against self-driving cars being viable.

If the median answer is $0 and there were a lot of negative answers then that means there's a lot of positive answers. Just sell the cars to those people!

Not everybody wants a self-driving car just like not everybody wants a big mac. Hell you probably need to pay some people to eat a big mac. But theres enough people that want a big mac that despite that fact it's a huge business.

Nonetheless, I would claim it's a lot more than half of people who would want self driving cars. How many daily commuters exist? I surmise atleast 90% of them would want self driving cars if it were a proven, familiar technology with a reasonable barrier to entry.

Of course, I'm just projecting a hypothetical technology that is both fairly cheap and extremely safe/effective. We can expect that future is far, far away

> Not everybody wants a self-driving car just like not everybody wants a big mac.

I don't think this will work, because self-driving cars will reveal drivers for what they are: impatient, reckless, anti-social maniacs. People want to speed, to swerve into the bike or parking lane to overtake a left-hand turner, to inch out into traffic to force themselves into a lane, to tailgate, etc, etc. You can't really have a system where x% of the cars are calmly and mechanically following the rules and manually-operated vehicles are taking advantage of that.

>The truth is, if full self driving cars actually happened, with safety statistics better than humans and with a low barrier to entry, you can bet that people would change their opinion real fast.

A lot of danger of motor vehicles is when motorist choose to be dangerous though. If self-driving cars force motorists to follow speed limits and other traffic control devices, that will limit their uptake by quite a bit I think

The average American will never be persuaded by statistics. Ross Perot tried that already.

It’s not statistics that will convince people it’s convenience. I just mentioned statistics because if the tech were convenient but actually killed people at 10X the rate of human driving, it could end up banned or something

It will be more like statistics convince some people, and then "my friend has one and he loves it, and I've ridden in it, so now I'm comfortable with it" will over time convince the rest.

It isn't the average American that needs to be convinced, it is the government car safety agencies (every government around the world with a such an agency, not just US). They have people who know how to read statistics. Once the statistics show significant improvements they will make anything else illegal. Of course significant is something that will be debated, and there will be a phase in time. In the end though, this isn't about the average person it is about the people with power.

Every time I read something like "are many decades away" and "less than one decade away" I ask myself if we have a prophet here... Since nobody knows what will be discovered today and tomorrow nobody can make claims that "something is many decades away". And on the other side, nobody can make claims for "less than one decade away", because some lunatic with the atomic bomb can stop the progress of an almost finished project... We can only speak about "probabilities" here...

>We can only speak about "probabilities" here...

Yeah but this is colloquial speech, the probabilities are there just implicit because it would be exhausting to make this explicit in everyday speech.

Investment is nothing but a game of probabilities.

If it's true that self-driving is likely to be developed in the next 10 years than Tesla and Uber are good investments, because there's a reasonable chance of a world-changing, market-monopolising product making your investment incredibly profitable.

If self-driving is unlikely in the near-term then Tesla and Uber are not so attractive - you're looking at investing in an overvalued electric car manufacturer with limited production and a money-pit taxi operator, respectively. If a moon-shot innovation like self-driving is still likely in the next 25 years then on the other hand there's not a whole lot of reason to believe that these companies are the one's that will get to capitalise on it, rather than the established (cheaper) manufacturers or some new upstart. There are far safer investments out there.

All that's happened here is that people have lost faith that these companies are likely enough to develop a world-changing technology to justify their recent high stock prices.

What about them is market-monopolising? Cruise and Waymo both are working on self driving cars and making progress (those are ones I know, I'm sure there are others that I'm not aware of). The traditional taxi businesses are watching this, and won't sit still.

> The idea that services like Uber without the drivers would make car ownership obsolete drove those valuations.

Did investors really believe that? I don't know a single person who believed that years ago when the pre-IPO valuations were going through the roof. I maybe knew a few people who believed it was possible one day but not anytime soon. It's wild to me that so called sophisticated investors really believed that in the not too distant future uber would have a fleet of driverless vehicles shuttling people around.

FWIW, I don't believe we'll see anything like that in my lifetime (I'm 37). The last few years of development in this space, for me, have proven that it's a much harder problem than people thought and that the early big improvements in the tech gave people false hope.

Supposedly, in some places, transport planners, who really should've known better, made decisions about mass transit projects on the basis that self-driving cars were coming any day now. If _they_ got it so wrong, why would investors do any better?

There was a weird period a while back where basically everyone except people who actually worked in the area seemed to believe the self-driving car was nigh.

My recollection of those cases is that those who believed mass transit was a boondoggle that no one uses were using the imminence of self-driving cars to argue against mass transit projects that they already opposed.

Most of those in charge of transit do not believe in transit. (in the US). Most politicians see transit as a way to shovel money into some special interest: they appoint a supported to lead transit as a way to reward help with no concern if they appoint someone who will be good.

>Did investors really believe that?

They didn't. I don't know what OP is talking about. Uber's self-driving initiatives were always a moonshot.

You say that like Musk didn't say a whole litany of things, including "By 2019, it will be financially irresponsible NOT to own a Tesla, because it will be paying for itself and making you money in your sleep as a Robotaxi."

Yeah - Musk says a lot of idiotic things, but even with that I maintain that these kinds of statements is not what drove Tesla (crazy) valuation.

I don't know who actually believed it, or if it had any real impact on investors, but it seemed to be a common comment on self driving threads on HN ~5 years ago that instead of ownership, there would just be idle fleets of cars that you would summon from, and you'd simply subscribe to a vehicle service as a monthly cost.

I still see that all the time. I don't think it will happen: there are too many advantages to owning your own car.

Perhaps some people did give a negative answer, but that isn't really why the median is zero. If you asked 100 people, and 51 said $0, and the other 49 said anything else at all (including all saying $1m) then the median would be $0. No negative answers needed.

> Uber, were priced as if self driving cars are almost here

Source? I struggle to even find it mentioned in Uber’s S-1 [1]. (To say nothing of Lyft.)

[1] https://d18rn0p25nwr6d.cloudfront.net/CIK-0001543151/f0dcd9a...

For so many years, the car ads were selling us freedom. The open road, and you're in the driver's seat. Driving could be fun, not just a means of getting somewhere. Sure, they were trying to manipulate public perception, but they were doing so in a way that resonated with a lot of people.

A robot driving your car? Where's the fun in that?

Look close. They are selling Dad (as in a male, and probably white) on driving, while they are selling mom (female) and the kids on enjoying the ride. As ads become less sexist you will see more females in the driver seat, but most ads are still selling the whole family experience, and that includes getting places while someone else in in control. Even when they do show females driving, the focus on the ads is less on the experience of driving and more on other features of the car.

that's a cherry on top. 99% of people come to unfortunate realization that in this country they need to buy some car to get their ass to work or daycare on time. That's what drives the big number. Everything else is just "now that you're here, let us show you how much more fun than the other brands we are"

I never understood how self-driving could not have been seen as race to bottom. At least if there was more than one platform in competition. It would happen in both selling vehicles and in offering their services. There is enough people that are very price sensitive and would go for cheapest option. Just look at airline market.

"Now that self driving cars are many decades away, instead of less than one decade away"

The promise was, that it will be here today. And this is clearly not happening and this is (one reason) why people feel bullshitted and so the trust and the stock go down.

So I also would not bet, that in 10 years FSD is ready, but I certainly see it as a possibility.

> Now that self driving cars are many decades away

So when I see these driverless cars in my neighborhood with no one in the driver's seat, is that just my imagination?

>The idea that services like Uber without the drivers would make car ownership obsolete drove those valuations.

Huh? What? No. That's not where the valuation came from. To be fair to Uber, they created a new type of market for which there was an immense demand for. Their self-driving initiatives were a moonshot and maybe got some people excited and therefore invested in Uber (see Tesla/Musk), but not quite the same as saying it drove their valuation.

>Tesla, ... were priced as if self driving cars are almost here.

Tesla was/is a hype stock based on the personality cult of Elon Musk that stemmed from being first-to-market with EVs. The valuation wasn't really driven by whether or not Tesla cars are self-driving.

You must have missed part of the huge hype bubble when Musk claimed that Teslas with FSD would be "appreciating assets" because owners would be able to rent them out as robotaxis when not using them. He said they should be worth up to $200k because of that, and also this p2p robotaxi fleet would make Uber and all taxis obsolete. If FSD was actually full self-driving, then the idea makes a lot of sense.

Yes, Tesla customers and investors believed it because of the personality cult of Musk, but there was an actual (albeit wrong and likely deceptive) business proposition that gave investors a reason to value it as being worth more than all the other auto makers put together. That proposition depended entirely on FSD being only X years away and that Tesla would produce the best FSD product.

https://electrek.co/2019/07/16/tesla-cars-worth-100k-200k-fu...

> You must have missed part of the huge hype bubble when Musk claimed that Teslas with FSD would be "appreciating assets" because owners would be able to rent them out as robotaxis when not using them.

If Musk actually believed this it was irresponsible of him to sell any cars at all since holding onto them would have yielded a better return for investors/stockholders.

Probably not actually. If you rent your car you, you also need to clean it. Musk/Tesla isn't setup to do that, and the overhead of getting setup for that type of operation means that while is a profitable small business idea it may not work out for a large company.

While there is a business in that space to be had, I'm not convinced it is very big.

That was the rationale for why they planned on not actually allowing leases of the Model 3 to buy out their lease at the end. Since robotaxi was right around the corner...

Not sure what ever happened there, but I think Tesla managed to hold firm on that stance and won by luck since the used market bubble happened...

> If Musk actually believed this it was irresponsible of him to sell any cars at all since holding onto them would have yielded a better return for investors/stockholders.

When Musk announced this, he said they were still selling Teslas because they needed just a bit more real-world data to 'complete' FSD.

Hype is not a crime. It enabled Tesla to access cheap capital. But Tesla went too far and starting collecting money for something that won't work anytime soon and possibly not within the lifespan of Teslas on the road now.

The promise was "buy this car and soon it will self-drive" (and for Uber it was "Never mind we don't have a sustainable business model now, we will soon with self-driving").

>The promise was "buy this car and soon it will self-drive" (and for Uber it was "Never mind we don't have a sustainable business model now, we will soon with self-driving").

OK - you can make that argument. The point I'm disputing is OP's statement that "Tesla, and even more Uber, were priced as if self driving cars are almost here." I don't believe that statement. That is, I don't believe that the valuation of either company was based on the idea that FSD was almost here.

Reply on news.ycombinator.com