Nevada approves regulations for self-driving cars | Hacker News Reader

Nevada approves regulations for self-driving cars

physorg.com

258 pointsdantheman133 comments

I'm very excited to here this because it solves the regulatory chicken and egg. Nevada gets a bump for being first (if things go well) and other municipalities can point to Nevada to craft their own regulations.

Kudos to Nevada on this one.

Yes! I hope this will snowball!

This makes me want to visit Las Vegas for this reason alone. I would love to go to Vegas and for the first time in my life ride in a driverless car, convertable, top down, and enjoy the view, take pictures, and relax no matter what the traffic is like.

I can see this being a MASSIVE boost to tourist cities.

I think there's only one street in the whole state that gets bad traffic.

Hey now! There are at least 2 other streets that get traffic at least annually.

It doesn't matter, traffic or not, I don't have to worry about driving, or street names, or traffic lights, or GPS directions. It's like the convenience of mass transit with the convenience of a private vehicle merged into one.

The bonus in vegas is more that you can be drunk and not have to worry about driving than it is the traffic. Of course, you could also just take a taxi.

Of course, you could also just take a taxi.

Oh it will be a taxi all right, just not driven by a human...

Actually, thanks to all the drunk drivers, badly designed roads, people from out of state, etc., traffic in the LV metro area is pretty horrible. Luckily there's a monorail on the strip.

s/to here/to hear/

This opens lots of cool possibilities such as eliminating the need for two cars in the family for example. You go to work, the car then comes back and drives the wife and then later in the afternoon when both parents are working it goes to school and picks up the kids!

It could eliminate the need for having a car altogether. Hello zipcar

Oh, yes. Having a robot taxi pick you up when you need it and disappear when you're done beats finding a parking space any day of the week. +1 city dwellers.

I'd like to be able to give my kids an RFID tag so that I could send them to school when I'm running late, or pick them up after sports practice. Trust will be an issue... but entirely doable.

zip car here in LA is way more expensive than owing an old luxury car AND paying dealership prices on maintenance.

or renting a car every other day. I paid 1/8 of zip car's prices when i used enterprise rentals for two months

Zip Car works best in SF, there are cars everywhere, the city is 2nd densest in the country but still a pain to do most errands without a car, cabs suck, and the price is right. Boston is similar, but I have not used ZipCar there.

In NYC it's at least $2/hr more expensive. Cabs become competitive, and there are tons of cabs. There are many more people but not as many cars as SF, so it's harder to get a car. It's mostly good for moving apartments without dealing with U-Haul, going to IKEA/Fairway in Red Hook or picking up craigslist stuff.

I found in LA, the zip cars were so spread out, I first had to figure out how to get someone to drive me 15 minutes to the car.

exactly what i was saying :)

to go the nearest market here you already spent mileage enough to get the rental for the day with unlimited miles.

and there's more car rental places than zip cars.

The downside is that cars could be driving a lot more. Now, instead of a roundtrips each for me and my wife, it is two for me and two for my wife (assuming worst case).

Individually, the costs go down (one car, woot!). Societally, (assuming anthromorphic global warming ;) the cost is much higher.

Of course, with one car, I can afford a far more green car, so maybe it balances out again.

the energy and material costs of producing a second car is very high. probably higher than the extra driving for the single car in the scenario .

Also, having a parking place at work for both cars (husband and wife) is also very costly.

Or your car could drive to a high voltage charging station downtown to recharge. When you need to do more than daily commuting you just rent a normal car.

First, I assume by the time that self-driving cars are mass produced we will also have affordable fully-electric cars. Second, I think you mean anthropogenic not anthromorphic ;)...probably.

Or if everyone moves towards self-driving taxi's, then the car takes you to work, then takes someone from near there somewhere else, etc, so you're really talking not much more extra driving, but saving tons on building fewer cars, fewer parking lots, and economies of scales for buying gas and maintaining the cars.

That won't help. Everyone in the area will likely be travelling to work around the same time.

Cars can be programmed to avoid traffic jams.

In some cities (including Washington DC) the public transit systems offer incentives for staggering your commute times. (Peak/Peak of the Peak/Off-peak fares).

I imagine something similar would crop up with self-driving cars to alleviate the rush hour capacity issues somewhat.

A setup like this should theoretically be able to optimize carpooling so as to minimize that. A smart automated system should be able to figure out how to efficiently cram 3-5 people into every car by optimizing routes.

Of course, in actuality there would be plenty of people who would demand their own private car out of principle.

Let the market sort it out, charge them more.

I think we should get rid of the 9-5 and have staggered times. Different companies start different times of the day... I'm fairly sure theres a group of who like to start at 10, and go home 6, or maybe start at 11.

> I think we should get rid of the 9-5 and have staggered times. Different companies start different times of the day... I'm fairly sure theres a group of who like to start at 10, and go home 6, or maybe start at 11.

It's worth remembering that companies are free to do this already. Most don't, however, because they want to maximise the amount of overlap with other businesses they need to interact with. It also conveniently fits in with other things in family life, such as the kids going to school, or sports, etc.

Businesses are able to afford this luxury of starting exactly when they wish because they are able to externalize the negatives of it... forcing the employees and road providing governments to take up the slack and suffer the costs. If everyone is taking self-driving taxis, it becomes easier for employees to negotiate that travel expense into the pocket of the employer. Costs should be borne by those making the decisions, else the invisible hand can not do productive work and this economic system has much less rational basis for being the one we choose.

Self-driving minibuses?

My guess is that we will have a very efficient Taxi system. Using a Smart Phone you will be able to request a Taxi where ever you are; it will already know your favorite locations; your favorite music; it will automatically charge you; we will not need so much parking space.

Sounds like a YC company in the making! In all seriousness, there will be some great business opportunities here. It will open up an entirely new market.

I always thougt companies like Uber were really positioning themselves and building the tech necessary for when this inevitably happens.

Right now they need humans to drive their cars around but as soon as they can they will surely shift to automated cars. If they've already built the scheduling systems necessary they'll be able to combine that with the mindshare they've been building and be in a great position.

Wow. I never thought of that, but you're absolutely right.

Uber is Netflix in 1997, sending DVDs through the mail to people's mailboxes, carefully awaiting the day when the technological infrastructure will support their true purpose.

Something tells me there's a subtext like that to a large number of YC-funded companies. They're clever.

It is probably a bit hard for a startup at this point, unless they just develop and license tech. The major cost and barrier to market is probably still in getting regulations through, once these are through though I think there will be lots of startup opportunities.

Yes, but will it automatically clean up the mess left by the scumbags who were inevitably in the car before you?

Your credit card, along with a few before, during, and after pictures of you in the cab should solve that problem.

As well as being linked to the person before you (for fines or something), I imagine you'd be able to send for another car, saying this one is too dirty etc. At which point it would be routed back to a facility of some sort.

Combination of financial and social prevention, automated surveillance and market segmentation will (relatively easily) take care of this. I don't see how it could be a showstopper.

I just hope that this goes into effect in a way that you would signal how many people will be traveling together and they send the appropriate sized car. Having every vehicle on the road seat 5 people and luggage when it is rarely at full capacity is extremely inefficient.

Riding alone? If so, they send a 1 person vehicle to pick you up. Riding with a friend? Two person vehicle. etc.

Launch app; app determines location; counter appears asking how many passengers (up/down arrows); push confirm, vehicle dispatched.

Agree 100%.

In fact, I think self-driving cars together with decent electric engines (which already exist) can revolutionize urban landscape.

Just imagine how quiet and pollution free our cities will be if the only traffic allowed inside are electric cars, who automatically go to charge themselves when necessary.

This could work incredibly well especially in urban areas. At least here in Europe, more and more cities are experimenting with traffic limitations in the inner city core, but this also has big drawbacks. Coupled with a cheap self-driving electric cars service, on the other hand...

Economically that seems like a wonderful idea, but politically it sounds doomed to failure. Taxi drivers tend to be pretty well organized politically, and in most cities even have laws in place (usually medallion systems) to limit competition. If they can keep these laws in place with their current very weak justifications, I wouldn't be surprised if they can shut down automated cars very easily, since the public would probably be a bit nervous about the idea anyways.

EDIT: Possibly I'm being too pessimistic. I expect automated cars could be really, really awesome and if so, and if they're allowed anywhere, eventually there'll be pressure in other cities to allow them too.

One of my concerns about an automated taxi system is that they'll become dirty and decrepit even faster than public transport because there's an increased level of privacy and nobody to maintain the vehicle between riders. How would you enjoy running late for something and the taxi pulls up, its seats covered in vomit? Eventually, all cars may be scratched up and 'tagged'.

Maybe this won't be as widespread a problem as I think it could be. Zipcar seems to be rather successful, although I've never used one.

In many cities there's a distinction between medallion taxis—which you can hail off the street—and livery service—which you need to call for. Since you generally have to take whatever you get get when you hail a cab, the prices are regulated and, in return, the competition is limited.

Generally, the number of livery service drivers and cars is not limited.

That said, the Washington, DC taxi commission has been hassling the Internet-based Uber. Last month they ran a sting against an Uber-arranged driver and impounded his car.[1]

[1] http://www.washingtonpost.com/blogs/mike-debonis/post/uber-c...

This is a beautiful example of a state, in desperate need of revenue, supporting legislation that encourages businesses to start up without increasing/decreasing taxes.

Now if only the US could do the same thing with stem cells.

Is the Nevads government subsidizing research into self driving cars with tax money? If not, than I see no parallel to stem cells.

For the last 10 years or more, you're welcome to do anything you want with stem cells. Go crazy. You just haven't been able to get tax money to fund it, if the stem cells happened to be embryonic.

There's not really an advantage to embryonic stem cells anyways.

Are you sure? How was this determined without doing experimentation with embryonic stem cells?

I thought the only 'difference' was the timeframe of availability, but apparently not:

http://stemcells.nih.gov/info/basics/basics5.asp

One major difference between adult and embryonic stem cells is their different abilities in the number and type of differentiated cell types they can become.

Scientists believe that tissues derived from embryonic and adult stem cells may differ in the likelihood of being rejected after transplantation.

They have actually managed to make adult stem cells pluripotent (research at the UW Madison did this), and furthermore have managed to make them from a person's own skin (again, UW Madison ftw). And adult stem cells are less likely to be rejected. Embryonic stem cells grow at a different rate than adult stem cells (they grow much faster) which causes higher rejection rate and can in fact act much like a tumor.

The difficulty isn't being able to research them. There's a lot of money flowing both ways that is very politically motivated, but it's pretty simple to say that the adult stem cells have actually resulted in recoveries and cures (embryonic have not) and do not have the ethical problems. Being able to research embryonic stem cells doesn't make them more viable. At this point, they don't have advantages.

wow, I didn't know that! Thanks for the education. Maybe I can fill in other industries (like the transcanada oil pipeline!) instead.

In an ideal world (according to economists), this isn't an issue. In the world we happen to live in, basic health research is publicly funded. Maybe it shouldn't be, and research funding should be purely commercial, or philanthropic. But with public funding crowding out commercial and philanthropic funding, banning public funding for some kinds of research has a similar impact to banning it.

How can public funding 'crowd out' other funding? There are more 'fund takers' than there is funding available, so if there were private funding available, it would be put to use. I don't understand your reasoning.

Research funding requires scientists. There's a relatively fixed supply of people who want to do research. Those who can get government or university jobs. Those who can't get jobs in the private sector. If you can't find any scientists, you don't employ any in your research lab.

This is an empirical observation, and isn't based on micro-economics, though you could frame it as such if you wanted. It's not true in all cases - Google might lure a few top computer scientists away from academia by giving them massive amounts of funding. See, academia has non-monetary rewards - the glory of working for as an academic (this may be a con), and government jobs look stable and for the public good (though libertarians might argue). Thus public funding secures the scares resource (talented scientists) at a lower cost, and private funding goes down (supply and demand, unless research happens to be a Giffen good, which I doubt).

If a driverless car cuts me off, to whom do I direct my middle finger?

Seriously, though, good on them - I'd love to see this kind of thing during my lifetime. Hopefully it goes well - I'd hate to see this cause a rash of accidents no matter whose fault it ends up being.

Whose fault?

SEGFAULT

If the accident involves a bus, does that make it a SIGBUS?

If something goes wonky and it hits a lamppost, SIGPOLL?

How do I send SIGSTOP to this horrible line of UNIX jokes?

Did you try turning your car off and then back on again?

With all of the sensors required, there will be a fantastically detailed record of every crash. My money is on the humans screwing up so badly the driverless car can't recover.

I would not be surprised if the first couple accidents were the result of nearby drivers 'ooh'-ing and 'ahh'-ing at the driverless car.

“It won’t truly be an autonomous vehicle, until you instruct it to drive to work and it heads to the beach instead.”

An oldie but goodie: http://www.computerjokes.net/027.asp

The Governor's comment about red license plates being used, then green once the technology is ready for the general public reminds me of the old regulations that were put in place when automobiles were just beginning to spread. Anyone driving an automobile was required to hire a runner to run in front of them and warn everyone to get out of the way!

I'm interested to find out how this will affect car insurance in the future. Although the human operator of the vehicle is held liable for any accidents, even if he/she is not physically present in the car [1], the likelihood of an autonomous vehicle accident will obviously be largely dependent on the car's software rather than the human responsible.

Will insurance companies stop charging you based on your driving history and instead determine your insurance rate based on the safety rating of the AI powering your vehicle? Or perhaps some combination of both? Insurance pricing based on AI safety could become a powerful motivator for developers of the software to continue to innovate and ensure driving safety.

There is certainly a lot of marketing potential in a company being able to say, "Not only is our car the safest autonomous vehicle on the road, but you'll also pay the lowest insurance rate as a result!"

[1] http://www.msnbc.msn.com/id/46419453/ns/business-autos/#.Tz2...

Could only really work if ALL cars are autonomous. How many car accidents involve only 1 car?

Also insurance is hardly an innovative industry... I'd bet they don't change anything about how they insure people for a LONG time.

Where's the link to the actual regulations?

I hate articles like this that could easily refer to the primary source, but fail to do so.

Does this mean we can also legally play poker with prostitutes in the self-driving cars?

Hunter S. Thompson would have loved this.

Am I the only one a little bit concerned by this? Don't get me wrong, I can see all the positives to come out of driverless cars, and the tech behind it astounds me.

My issue is I love driving, and once the world sees how much safer the world is without driver error, would driving manually eventually be banned?

I realise I'm essentially putting my enjoyment over people's safety, but what can I say, I love driving. Just have to hope riderless motorcycles don't come along anytime soon.

Having said all that, it's just occurred to me that if all cars we're driverless, then motorcycle crashes would be massively reduced. Maybe banning manual driving wouldn't be so bad.

>My issue is I love driving, and once the world sees how much safer the world is without driver error, would driving manually eventually be banned?

I think it'd be more likely to be moved somewhere off-road: Race-tracks, specific off-road areas, that kind of thing.

Continuing to allow the day-to-day wholesale slaughter that is humans-driving-cars on public roads is - frankly - insane.

(edit: for slightly better grammatical clarity)

If you look at how current safety systems are designed, there will probably be a manual mode with the car taking over if it thinks your doing something stupid. There will probably always be places where you can drive cars built without such systems.

That said, if you want to drive a 3,000+ LB vehicle at ridiculous speeds and or in a completely unsafe manor on public highways, Fuck You no. Your desire to take stupid risks for fun ends when it puts others at significant risk, you sociopath.

I wonder how long it will take until it is illegal to have a human driving a car. I am sure self-driving cars will cause much fewer accidents. Even better: once all human-driven cars are gone, traffic can be organized incredibly efficiently, esp in areas without human interference like on Freeways. Cars won't need to stop anymore. Traffic lights could communicate when they will be red or green well in advance and allow cars to slow down or speed up a little. That would greatly increase fuel efficiency and allows prevention of traffic jams.

Brian Hayes had an interesting column last year on what could happen if these became common, or even universal.[1] One advantage he pointed out was that you wouldn't have to waste time looking for a parking place. The car could drop you off right at the door and then go, maybe far away, to park. Valet parking for everyone!

[1] http://www.americanscientist.org/issues/id.13359,y.2011,no.5...

The question of the self-driving car is much more complicated I think than the benefit it will bring. The consequences to the labor market will be interesting to say the least. Just imagine.

No taxi drivers

No UPS delivery truck drivers (automated delivery anyone)

No truckers

Less cars produced, less people building cars.

Less doctors, hospitals etc due to less accidents.

I don't remember who said that we are entering a period of massive labor change. A second industrial age where there is less need for labor.

The first one brought marxism and communism what will this one bring?

True socialism is my guess.

Socialism is impossible, due to the inability to use economic calculation when the means of production are State controlled.

When you don't have profits, and you don't have market prices, then you have no idea when you should be switching away from some resources and using others. Resources get squandered, there's no way to make sure they go to their most valued use. That's the beauty of Capitalism, we make sure resources go to where they are most needed by selling it to the highest bidder.

Ex: If copper prices go up, some people - those who can - will substitute away from it, thus allowing the copper to go to it's most valued use.

If interested, I recommend 'Economic Calculation in the Socialist Commonwealth' by Ludvig Von Mises [http://mises.org/econcalc.asp]

Also good is Friedrich Hayek's 'Competition as a Discovery Procedure' [http://mises.org/journals/qjae/pdf/qjae5_3_3.pdf]

You do know there are socialist models that also use markets?

Don't forget a huge amount of space reclaimed on the roads and in city centre parking lots. Most cars are sitting idle for 90+% of every day. With self-driving cars, you need 1/10 the number of cars to achieve the same result (maybe 2/10 to deal with rush hours).

It's incorrect to say "less need for labor" because labor in general is the most scarce resource (it's easy to prove: we have unused natural resources and capital always requires labor to make them work and maintain them). This is exactly why everybody tries to build capital which makes labor more efficient.

It is clearly more efficient for a man to control 10 cars with a computer instead of the one with his bare hands.

What other 9 persons would do? Either control more cars (if there rises a demand for them) or shift to remaining, more important processes where the labor is still less efficient (and thus costs more), or any combination of these. Think of it like "we solved the problem with cars, lets now move on the next task".

And as you know, existing capital and technology lets people learn and adapt more and more efficiently. So it's not "progressing economy needs less labor", but rather "progressing economy makes labor more efficient by solving some productivity problems and shifting labor where unsolved, more urgent problems remain."

True, but those unsolved, more urgent problems often require more educated labor. Many taxi drivers who find themselves out of a job likely don't have the skills necessary to quickly reintegrate into the labor market solving the tougher problems you're talking about.

As machines continue to replace humans in tasks that can be automated, fewer unskilled jobs will be left for less-educated segments of the labor force. As technology automates new processes that previously required human labor, the educational barrier to entry becomes higher for the jobs that are left, because the easier productivity problems have been solved.

Yes indeed. This is why I'm so interested in the educational revolution that is currently happening. Education is important in this situation.

The good thing is that the higher end jobs tend to be creative(Research), and in near infinite supply if you can get money flowing into them. You can never have enough inventions or new technologies.

Is this a serious argument? One could say the same thing about:

1) The advent of modern gardening and genetically modified crops. With increased efficiency, where will the extra farmers go?

2) The advent of computers. What will happen to books? The tree cutters? The people who ship the trees to the lumber plant?

I could go on and on about ANY technological advance in the last 200 years. I don't see how cars are any different that will bring "marxism or communism".

I have similar feeling. Heck, invention of farming have caused many the hunters and gatherers to lose their jobs!

I can't believe how "taking away jobs" sort of argument against technological advancement keeps showing up over and over, despite being refuted every time.

My favorite comeback is, "Think of all those poor switchboard operators that the Internet put out on the street."

Eh, it might be right one of those times. People have been retreating up the skill chain as technology has become more and more advanced. If we could make a tool that does everything as well as humans for much less, society would have to change massively. It has in the past for lesser changes in technology.

>If we could make a tool that does everything as well as humans

That goes so far beyond anything we have any experience with that it's impossible to speculate. Every technological advance in history has been a labor saving device; a device that does everything as well as humans is no longer a tool, it's our replacement (i.e., strong AI).

>People have been retreating up the skill chain as technology has become more and more advanced.

The majority of jobs created by industrialization required less skill than previous jobs (e.g., farmers to assembly line workers).

Of course, I'm just saying that the limit of technological advancement as t approaches infinite in some scenarios implies no human employment, so it's not obvious that just because permanent mass unemployment has never happened that it will never happen.

Since industrialization, the skill required by good jobs not made artificially lucrative by unions has become more specialized and advanced. Whereas being good at manual labor has become less useful. That's what I meant by going up the skill chain - it's harder to replace thinking jobs with tech, so more recently, people have been trying to work in those.

Usually the savings from employing these people, will end up being spent somewhere else(More eating out for example?). Allowing a different sector to grow to compensate. The money being spent on their salaries does not disappear... and has to spent or invested somewhere.

Been in contact with politicians in the Swedish Parliament trying to alert them to this, but they don't really seem to care, I even got "Well, I like driving" as a response from one politician. It's sad, given that Sweden is a country where a lot of cars get tested that we're not seizing the opportunity to allow for innovation to happen in the car industry.

Is it possible that this will turn cities in to less desirable places to live? Imagine if you could take your car in to the city for drinks, have it drive around for a while and pick you up when your done.

That is just one possibility, I'm sure there are so many applications I can't even think of them.

maybe the opposite is true. i can get around the city very fast without having to worry about parking.

I think this is more likely. Living in cities is desirable because you have a vibrant and 'close' community, it is undesirable because its hard to drive and if you do own a car its under utilized.

One of the interesting effects on this might be 'non-bus' cars. Which is to say public transit as a fleet of say 'four person' self driving cars where you walk to a bus top and touch your transit car and one shows up and picks you up. You tell it where you want to go and it takes you there.

It of course destroys the taxi business quite completely. But combines all manner of public transit into a single service which gets economies of scale. Dynamic electric cars would be fine (where they have an electric 'boom' like trains in Europe do).

No way it's public. I can imagine apartment buildings having a car or two for use like a gym or laundry.

Or, perhaps a subscription service, superzipcar or something like that.

Why "no way"? I don't have any problem with your suggested models either, but see no reason why public car services couldn't coexist.

short version, politics & tragedy of the commons.

Anything is possible, I just can't imagine a (US) municipality that's able to implement an entirely new public service. IIRC, community wi-fi keeps getting sued by companies that want to make money. First, none can afford it. Second the US, IMHO, is really hostile toward any sort of government service.

Even if those were non-issues, you still have to solve the problem of vandalism. People behave like grownups when other people are watching (pretty much). The bus or subway are good examples. Alone, in a car, i think it's much tougher to keep people well behaved. there have been various community bicycle programs that suffer from people beating the hell out of the bikes, or just stealing them.

These are perhaps mutable problems in small towns, or outside of the US.

Seattle has a light rail system they only booted up a few years ago. They had to build a completely new set of rails everywhere to do it. I'll admit I don't know the details of its history, since I wasn't paying attention, but it's utterly possible.

If you belive this (http://www.thetransportpolitic.com/2009/07/20/seattles-light...) it took seattle almost 20 years to make that happen. Also, i think that city, in particular, was flush with cash.

20 years is on the long side, but I find it a reasonable amount of time to make it happen. And the point was more that it happened, and that it wasn't exactly an isolated case.

It doesn't need to drive around for a while. There could be parking lots for driverless cars that somehow connect with the owners credit card and pay for parking.

I'd be surprised if cities became less desirable. I could see cars turning into something like an on-demand, private bus (or just an automated taxi that you own).

The proliferation of self-driving cars will absolutely affect where people live, eventually.

I imagine that to start with, we'll see certain areas - such as the Vegas strip, perhaps - that support Valet™, a feature first appearing in luxury vehicles that does exactly what we'd already expect from human valet drivers.

But to drill down even further, I'm not exactly sure where the very first Valet™ supported parking lot would be. Manhattan would be a great use-case, but my own guess is that we'll first see this appear at places like large mall parking lots where city regulations won't be as much of a problem.

And yes, self-driving cars will provide a huge boom to services like Zipcar, and indirectly to marketplaces like AirBnB that convert ownership into access.

At first I thought that taxi companies would use the vehicles, but now I think it is more realistic that zip car would be first (or someone like them) since not everyone will be fast to adopt and may be uncomfortable riding.in one. Then over time the economics will force cab companies to switch at which point it will be a mature market.

Bschool professors will talk about how cab companies made the wrong strategic choice even though it will really be the only option and one essentially predetermined.

Personally, I can't wait!

Something that is needed here and probably many places is competition in Public Transport. I imagine a time where it will be cheaper to have your electric car drive you into work than to catch a train, then drives itself home and charges until you're ready to be picked up again. Or just driverless taxis on the cheap.

Public transport here costs at least $20 a week just for students relatively close to the city. You can expect $50+ for adults, it's not uncommon for people here to live 1+ hours away and commute from the coast, costing $15+ per trip.

Regardless of whether or not it drives itself (although it's an improvement), I still find the idea of owning a car very unappealing, and it still doesn't give the suburbs a soul.

You wouldn't buy a car, you'd just order a ride when you needed it, and the car would come pick you up. Think Zipcar

But why did people leave cities to begin with? Street crime was an important factor.

It's a complex question, but I think the main reason was just for child-rearing space. Crime wasn't an issue when the first suburbs exploded in the 50s.

With fewer people having children I think that's less of an issue.

more details in this (slightly) older article: http://www.nytimes.com/2012/01/24/technology/googles-autonom...

There's no hope, all taxi guys will be unemployed. Not now, but soon. And this trend will continue, everyone, even programmers in the not so far future have to study and become architects.

America, if there's anything that's going to send bolts of lightning through your economy, it's this.

Unrelated, if driverless cars are efficient enough, then traffic lights won't be needed and traffic will merge continuously.

Does that mean pedestrians can confidently stride out to cross the road and the traffic will effortlessly stream around them?

So I assume Google is going to start selling ads and collecting information through autonomous cars?

Can't wait for the day my car will drive itself to the shop for a smog check and oil change.

Regulations is one thing, but when do we expect self-driving cars to reach the market ?

self driving cars in nevada can be accomplished by a timer.

Aim the car. accelerate. engage cruise control. start timer. After N minutes required to get to the place, stop car.

That's great!

Having just read Robopocalypse, I'm very concerned with this development. These automated cars did a lot of damage in that book.

just the legal question; if I drive with a family and autocar will smash into me, killing my children and wounding me so I cannot walk or work, who is going to be liable? car manufacturer, producer of the tires, software engineers house, or GPS system? Could see 20 years of litigation between all parties involved before I get a dime for recovery, or probably die first.

I've yet to read Nevada's regulations, but having it be the same as for regular cars would make the most sense. In an isolated accident, the operator's insurance company is liable. If there's a fundamental flaw with the cars, the manufacturer issues a recall. It's far more likely that an inattentive human driver will smash into a driverless car than the reverse, though.

I guess what I was trying to say is that cars were designed to be driven by humans, and although there are situations where machine's fault is hard to prove [1], in reality adding additional layer of decision-making piece of hardware will make it even harder to determine who's fault was when my children, hypothetically, died in a car accident smashed by a second car.

[1] http://www.dailymail.co.uk/news/article-1300880/Toyota-defen...

Almost all auto fatalities are caused by human error. I think saving 1 million+ lives per year is worth the legal wrangling to determine fault for the remaining few accidents.

There are already rules applied by insurance companies for deciding liability in an accident. Just apply those same rules as if your over-the-top rhetorically invoked children were killed by a human driver. Retributive criminal penalties for accidents should be done away with in favor of civil action handled by insurance companies, except in cases of gross negligence or willful harm (e.g. the owner of the car instructed the computer to cause an accident).

one possibility is today's system - whoever owns the car (whether it's a private individual, a cab company, or the city) will pay for insurance, and that insurance will pay out to cover any damages.

Now, what happens if there's a bug in the software, causing a lot of damage at once? Can the insurance sue the programmers to recover losses? Probably; but at the least, the victims should be paid promptly by the insurance contract, without the need to wait for a court settlement.

Yeah, but if insurers have to assume that liability until they can maybe recover damages (just think, the car manufacturer or software firm might not even exist anymore and the car still on the road), insurance is going to be very expensive for these things. Insurance companies aren't a charity, they're going to have to pass on those losses into the premiums.

I think the only way it works is if the state certifies the vehicle and requires a liability policy that will pay for any incident.

If the systems are good, I think such policies will be cheaper than liability insurance for a bad driver. If the systems are bad, I wonder such a policy will be available at all.

The manufacturer would still be on the hook for defects found after the certification, and I expect defective vehicles with no remaining manufacturer will simply not be licensed for operation.

I think one of the main issues with liability here isn't the expected loss - I would be extremely surprised if driverless cars weren't much safer than ones with drivers, at least once the major kinks are worked out - it's the correlation between accidents. Currently, accidents happen more or less at random; when controlled by software, a bunch of cars might go berserk at once, bankrupting an insurance company in a fit of bad zeroes and ones.

Automated hospital equipment might be a place to look for precedent in regulations, not sure what they are but it is a similar situation in that a machine malfunctioning could cause injury or death.

Reply on news.ycombinator.com