Why you have (probably) already bought your last car | Hacker News Reader

Why you have (probably) already bought your last car

bbc.com

22 pointsmooreds32 comments

So frustrating to me that otherwise smart people keep buying into this narrative, including software developers and tech people who should know better.

Here's the actual status quo: after a decade of research into autonomous cars specifically (not to mention the decades of AI and systems research before that), we are barely at the point where a car can function safely with Level 2 autonomy, in good conditions.

These prognostications about eliminating traditional cars are assuming reliable Level 5 autonomy in all conditions. And unlike most software, you can't just get 99% of the way there and call it done. If you want to remove the steering wheel, it has to work ALL the time (or close enough to make no difference.) They say that 80% of the work is in the final 20% of the project? We haven't even finished the first 30% of the work yet.

Even more than that, nobody seems to be talking about the fact that to get this to work reliably, you basically need full general intelligence (unless you can manage to turn the road system into a closed system, like rail is). For example: following verbal or gestured instructions from a police officer or construction worker. We don't even have _theories_ on how to build intelligence like that, yet.

Will it happen eventually? Sure, probably, eventually. Will we get a lot of use out of Level 2 and Level 3 autonomy in the mean time? Already happening. Will we be taking robotic taxis in the next 20 years? No way, not with current approaches absent a breakthrough in general AI.

Do these requirements for the levels you mention anticipate the changes in traffic patterns that will occur? What I mean is, if we could save all the space that the author is noting, could we relegate entire lanes of bikers into completely separate streets? If bikers were gone, for example, could we get away with not needing cars to go to level 5? I'm speaking hypothetically about one hazard but has anyone modeled what would be the environment for cars if everything changes? If your train travels in an enclosed tube, it would require fewer safeguards against obstacles, and could we put autonomous vehicles in a similar kind of enclosed environmental tube?

>For example: following verbal or gestured instructions from a police officer or construction worker. We don't even have _theories_ on how to build intelligence like that, yet.

The main problem with having humans take over control in situations the car can't deal with is if the switch needs to happen at speed. If you're crawling along near some accident getting input from someone directing traffic, it doesn't seem like it should be a problem to get the human's input with voice commands, or touch on a dash-mounted tablet, as to what to do next.

> Here's the actual status quo: after a decade of research into autonomous cars specifically (not to mention the decades of AI and systems research before that), we are barely at the point where a car can function safely with Level 2 autonomy, in good conditions.

It's even worse. I remember seeing prototypes of autonomous cars a little bit above 20 years ago. And I don't know when that research started.

> following verbal or gestured instructions from a police officer or construction worker.

If they don't broadcast it with a valid signature, just ignore them, they're not authorized to give any instructions.

> turn the road system into a closed system

You can turn it into a "closed enough" system, with every vehicle having a radio locator, and everything else being treated as an unknown obstacle... that someone could get fined for putting it on the road.

You've never encountered a citizen flagging down traffic because of an obstruction just past a blind hill or curve?

People are already taking robotic taxis:

https://www.theguardian.com/technology/2018/aug/28/driverles...

Admittedly, on a very limited basis. But there is one thing to consider: There are a lot of city cores where average speeds are low enough that such specialised services are viable way of reducing the need for fully flexibly autonomous cars.

E.g. here's an autonomous shuttle manufacturer:

http://www.easymile.com/

These are also in actual use some places (though lagging laws means that many places there's still an attendant), and there are other manufacturers.

So while the full general purpose replacement may take longer than hoped, a lot of more special purpose vehicles like that will nibble away at transport needs, and illustrates that if we look at it as not a problem of replacing cars, but a problem of replacing journeys, it's possible to solve parts of it piecemeal long before we can fully replace cars.

"(unless you can manage to turn the road system into a closed system, like rail is)"

I still don't understand why everyone is working so hard trying to make smart cars when it seems like smart roads that control dumb cars would be a simpler, if more expensive, solution. Why not make it a closed system?

Because it is open. Things like wildlife, children, drunks, falling trees, snow, ice, flood waters, rock slides, and unsecured cargo can enter the road at any point.

Honestly, I bought my last car a decade ago.

Cars aren't that great for a lot of reasons (my stress level, my city's livability, the environment), and getting rid of the idea of car ownership was one of those huge life decisions that I made that I've never, ever regretted.

I don't use car sharing services - that whole industry is toxic for the drivers, I sometimes take public transportation, and every few years, I take a flight.

Everything else is done by bike. I'm happy, I'm healthy, and I'm looking forward to the rest of my life without owning another car.

But where do you live?

Read the article in his profile, you'll find your answer in 10 secs. And yeah, with that profile, in that location, I take his point. But not everybody is that lucky :)

I'm increasingly starting to wonder if the autonomous cars will get the chance to "win" before being replaced by "flying cars". There's half a dozen companies at least that are aiming for first flight in the next 2-3 years, and Uber-like services in the 2025 timeframe. Some, like Lilium have demonstrated flying prototypes.

Companies like Lilium are claiming they'll be able to compete with and eventually beat Ubers current pricing, and fly you at speeds of 300km/h....

If that comes through, it will have a vastly more disruptive effect than autonomous cars. E.g. I live about 10 miles from the centre of London, and that's about a 1 hour commute. A one hour flight at 300km/h would make most of England viable to live and work in and still have commuting distances similar to what Londoners consider perfectly viable commuting time today...

It has the potential to change the dynamics of both the housing and labour market dramatically. And who knows what it will do to e.g. downtowns when most of the country suddenly is within 20-30 minutes or less of one of the half a dozen or so largest city centres.

To a European used to seeing American sprawl as an example of emphasising the car over public transport the prospect of something that can multiple acceptable travel distances by a large factor over the car is a bit scary...

Flying cars will always be less efficient (and thus, more expensive) than driving cars due to the physics of it. I wouldn't hold my breath on flying cars replacing driving cars, just as airplanes haven't replaced trains.

> Flying cars will always be less efficient (and thus, more expensive) than driving cars due to the physics of it.

As it stands, for cars, the typical energy use is 3k-4k BTU per passenger mile. Commercial airliners use 2k-3k BTU per passenger mile. Of course those get their efficiency from packing more passengers in. But the point being that a statement like the above only makes sense when comparing comparable vehicles (the bad numbers for cars is in part because cars on average have less than 2 people in them; fill a car up, and it will do better per person than a typical plane).

However the best reported numbers I've found from test flights of electric planes is energy use as low as below 1300 BTU per mile (these numbers will also vary wildly by number of people; most prototype models in progress are for 2 people, including Lilium, though their next version will be a five seater).

Part of this is due to totally different design constraints. Of the 12 in-progress models I'm aware of, the heaviest is about 1/4th the weight of the lightest Tesla. The lightest is about 1/6th the weight of the lightest Tesla. E.g. the two-person version of the Lilium was 400kg without passengers.

Part of this again is that the cost of purchase vs. cost of operation calculations are entirely different: Most of these are designed for the hire operator market or as luxury items, and they're not even possible if they weigh too much, so much more money gets spent on expensive materials to push weight down. Their assumption is that they'll get the operational costs down far enough to amortise the increased finance costs of buying it.

Yes. The Uber Elevare white paper looks at this issue, and the stylised numbers are:

Car: 1 mile/kWh

Flying car: 2 mile/kWh

Electric car: 3 mile/kWh

(Electric engines are much more efficient than combustion engines.)

So, it’s not too bad, plus ground based vehicles (according to the same study) have to drive on average 1.4 miles for 1 mile distance “as the crow flies”.

However, the numbers above are for fixed wing flight, as far as I can tell. Drone-like hovering flight is less efficient.

Still, I think that’s not the main problem. Regulation, battery specific energy, ATC are bigger hurdles.

https://www.uber.com/elevate.pdf/

We've had 'flying cars' for 100+ years already, and automobiles are still around. Flying just doesn't make sense (energy, safety, and practicality-wise) compared to rolling.

We haven't though. The distinction between these aircraft and typical planes being that they don't need a runway (several of them are VTOL), and are aiming to do away with the need for a pilot. The big limiting factor for small planes have been that combination of cost, inconvenience (nearest airfield to me is far enough away to make it pointless, and most places I'd want to go to don't have an airfield nearby either), and the need for a pilot.

First when you solve all three of those, you have something that is addressing the same market.

We will see whether or not it will make sense - there's now half a dozen manufacturers at least that claims to have electric aircraft in this class that are efficient enough to be comparable with electric cars in energy use. Of course it's possible they're all wrong.

> there's now half a dozen manufacturers at least that claims to have electric aircraft in this class that are efficient enough to be comparable with electric cars in energy use.

You'd have to use a very loose concept of “comparable” for such claims to be plausible, but the personal flying vehicle market has been rife with vaporware draining the wallets of investors and enthusiasts (the latter through early sales of reservations for “imminent” models) long before “electric” was part of the usual pitch.

> You'd have to use a very loose concept of “comparable” for such claims to be plausible

There are actual flying prototypes that have gotten at least within a factor of two of electric cars, depending on respective passenger loads. How they'll come out with production models remains to be seen, and also depends greatly on average passenger counts.

People tend to overestimate how much energy flight takes - consider that both large scale commercial jets and small private aircraft are relatively comparable to cars in energy use per passenger mile, often coming out better for larger planes than cars (though admittedly largely because average passenger numbers for cars is so low).

> but the personal flying vehicle market has been rife with vaporware draining the wallets of investors and enthusiasts (the latter through early sales of reservations for “imminent” models) long before “electric” was part of the usual pitch.

That's true, and I got a good laugh of e.g. seeing a pitch from a company in this space that included the Moller SkyCar as a competitor (I'm not sure if they were serious and included it for completeness, or if it was in there as a joke; they included the year it was announced, and that it was not yet available, I'm presuming it was a joke), but there are a few significant differences: The rapidly improving energy density for batteries coupled with much simpler designs and more advanced materials allowing for lighter planes. Add in autonomous operation and the cost picture for operations changes dramatically. This to me is the biggest change: Several of them are now aiming for "Uber in the skies" rather than selling to individuals, which allows for entirely different financial tradeoffs.

At this point there are also multiple flying prototypes, which is a marked difference.

> There are actual flying prototypes that have gotten at least within a factor of two of electric cars

Linky?

> People tend to overestimate how much energy flight takes

But underestimate how much energy convenient VTOL takes.

> Linky?

This is one of the easiest accessible ones: https://www.theatlantic.com/sponsored/thomson-reuters-why-20...

> During a 62-mile stretch of its historic flight, the plane used about 25 kilowatts of electricity for a total energy cost of just over $3, according to its pilot, Klaus Ohlmann. In all, the e-Genius consumed just a fifth of the energy of a typical, fuel-powered two-seater airplane.

I can't share the rest I have unfortunately, as while the information itself is public (I believe the source is a study by Deloitte), the document I have also contains information that is not, and I don't have a publicly available reference for the source. I'll see if I can find one.

> But underestimate how much energy convenient VTOL takes.

I don't think so. E.g. Lilium uses VTOL, but the wings (holding the electric engines) starts tilting as quickly is possible to start taking advantage of ground-effect, as everyone is of course aware that VTOL is expensive. While the power they need for that short period of time is higher than ideal, it still ends up being a negligible portion of total power consumption other than for extremely short hops.

One word, that I think/hope is becoming important to more people these days, "privacy" (not mentioned once in the article). Unless these things can be used without identification, I seriously hope they don't become the only, or even a common, transportation option.

With electronic license plate readers everywhere, you already lack privacy in a private vehicle. I wish there were a way one could drive anonymously yet still be held responsible for hit and runs, etc, but that doesn't exist.

Somewhat true, but is that data for sale on the open market? I mean, I wouldn't really be surprised if it was, but I'm hoping it isn't, at least in most places. I'm OK with government officials having access to data via a system of checks and balances, not so much with just anyone that happens to have a few bucks. Just because we've lost some privacy in one place, doesn't mean we should just stop caring, that's how things have gotten as bad as they are.

I hate to say this but.. There will never be autonomous cars. Human psychology, and unpredictability, will doom it all. We can build amazing AI technology, but we will never be able to account for Jenny after six mudslides in a Geo Tracker, or the random decision making of even the most well intention human driver.

Furthermore, if you haven't read "Industrial Society and Its Future"[0] you should. It makes a very good point that society as we know it today cannot and will not support a car-less reality.

[0]. http://editions-hache.com/essais/pdf/kaczynski2.pdf

As a layman, I think this title is way off, even if the internal concepts of the piece are reasonable. We're a long way from fleets of autonomous vehicles being ubiquitous enough that I can expect a vehicle of the specs I require to be available in the same amount of time as I could start one in my driveway. Maybe this isn't fully informed, but my impression is that even an early automobile was strictly better than a horse, for the riders in the moment.

That said, the average person may have bought their second-to-last car. But I bet you we end up owning passenger/cargo modules that dock at our houses for transport platforms to pick us up in.

As someone who lives without a car: you don't really need to be able to just start one. The big differentiator in my experience is whether or not the wait is consistent.

I will ditch a cab company if the waits become too variable, but will happily accept a few minutes longer wait, as it's not really a matter of whether it's 3 minutes or 5 or 10, but knowing that I need to think about it at least that much in advance. If I regularly go to book one 3 minutes in advance and 25% of the time it turns out to take 8, that's more annoying that if it's consistently a 10 minute wait.

As long as it's consistent, it becomes habit very quickly to order at the right time in advance.

That's also a great way to sum up the problem with many forms of public transportation in the US too. The light rail and buses near my home are reliably unreliable, sometimes they are right on schedule, sometimes they arrives 10-15 minutes after schedule, and sometimes they arrives somewhere in between.

The horse was autonomous driving with collision avoidance, voice commands, no steering wheel needed. But horses left a lot of dung in the roads and needed lots of pasture. So the horse, despite wonderful self-driving modes, had to go.

Currently we only have a few EV cars in the West, apart from Tesla and Nissan/Renault nobody is really up to speed with it, they are all still churning out these old-fashioned ICE cars and not making a lot of money at it. Right now Ford are pulling out of cars as they cannot do so profitably, they need to sell tank sized trucks that are a tax fiddle for them to make a profit in their domestic market. The Fords, GMs, FCAs and Daimlers of this world have left it too late.

Meanwhile, in China we have plenty of EV cars being actually made and driven, not talked about and waited for. The Chinese government fully expect to take over the world when it comes to making electric cars, buses and trade vehicles. I can't see Donald Trump with his tariffs (taxes on American consumers) turning this around.

Although Ford might see making a $20K car as impossible, even with an old-fashioned 'ecoboost' engine, never mind an electric one, companies in China might be more than willing to enter the low end of the market and be able to deliver a profit. Think of how the Japanese were able to do so with economical cars two generations ago and how Korean companies have came along too. They were happy starting out with the low-end that Ford, GM and Chrysler couldn't really do.

Meanwhile, people might realise they do not need 300 miles of range and that only 30 miles range is needed for what they actually do.

The less batteries you have the less batteries are needed to carry those batteries. So a city car could have only 30 miles of range and a nice little supercapacitor to give it some 'ludicrous' grunt.

Plug it in when at home and at the shops and it will always have those thirty miles, no range anxiety needed. Electricity could even be the new 'free parking'.

These chargers are coming, soon pubs will have them in their car parks and if they don't have them then they won't be getting any wealthy customers. So these things will just appear.

If you were to strip all the 'falcon wing doors', 'alcantara trim', little lights where you don't need them, bits of carpet in the boot, pointless speakers and other nonsense that you get in one of today's cars then that 30-mile range battery can be even smaller. A light and efficient EV will be more pleasant just due to being quiet, customers won't need the silly 'extras'.

Since most people don't buy cars but have some finance hack going on there is a high likelihood that the people that buy new cars will be buying new ones in the next 1-3 years. Next time round they might decide to go electric. If by then the Chinese armada has arrived and there are 'Dacia' grade EV cars that cost $20K then why would you want to pay for anything else? Of course the older generation who watch Jeremy Clarkson and like their 'vroom-vroom' noises will stick with ICE but those folks will not be around forever. The younger generation will soon have no interest in those things much like how nobody is that interested in long playing records these days.

Buyers in the second hand market might also decide to not have some old-person's hand-me-down ICE car and want something more practical such as an affordable Chinese EV. Why would you want to buy something that costs a fortune to be serviced? Might as well get a horse! So it should be a no-brainer to get an affordable EV.

The article shows New York with no horses a century ago. There were probably lots of people still using horses in the countryside. It will be the same with the EV switchover, in big cities like London (maybe not New York) there will be just EVs but in the countryside, e.g. rural Scotland, people will have the old-fashioned cars still.

So the article is far from way off. The autonomous driving 'gig economy' nonsense isn't coming for a long time, if at all, but practical EV 'motoring' is definitely possible in a short period of time. How long did it take for the world to go from dial-up to broadband? It is that type of timescale.

I believe car leasing will become more popular than buying as automobiles become more technology than machine. Who wants a gadget car with 3 year old technology? People can't even keep phones that long.

There is a less favorable reason you are right. Cars are getting more expensive and people can't afford them even with ever lengthening loan lengths...

Reply on news.ycombinator.com