Forum adverts like this one are shown to any user who is not logged in. Join us by filling out a tiny 3 field form and you will get your own, free, dakka user account which gives a good range of benefits to you: No adverts like this in the forums anymore. Times and dates in your local timezone. Full tracking of what you have read so you can skip to your first unread post, easily see what has changed since you last logged in, and easily see what is new at a glance. Email notifications for threads you want to watch closely. Being a part of the oldest wargaming community on the net. If you are already a member then feel free to login now.

Traffic jammers will become a viable reality.

Interesting question of ethics, actually. Should an autonomous car always stop for people? If yes, what keeps idiots from causing massive jams by keeping vehicles from moving at all just by dancing around between them?
If people can always just bumble into the streets without looking, city traffic will be ridiculously slow.

XuQishi wrote:

Traffic jammers will become a viable reality.

Interesting question of ethics, actually. Should an autonomous car always stop for people? If yes, what keeps idiots from causing massive jams by keeping vehicles from moving at all just by dancing around between them? .

In theory you can do that now with cars driven by real people. In fact people have done just that - normally when they do the police get involved. The only difference is that self driven cars might well react faster so there's less potential risk there; however chances are the self driven car will have all round video recording so if you do do that you will likely get caught even if you flee the scene before the police arrive.

. In fact people have done just that

Well, at least the robot doesn't get out of the car and tells you forcefully what he thinks of you .

XuQishi wrote:

. In fact people have done just that

Well, at least the robot doesn't get out of the car and tells you forcefully what he thinks of you .

Driver (ergo passenger since the car is doing the driving) gets out to forcefully put across their point of view and thoughts on the matter.

Obstructing person runs off

Driver gives chase for a little bit

Car detects there's no obstruction - drives itself away!


(ok so any sane person getting out in such a situation would turn the car off; but still its a humorous thought! Right up there with if the car has sensors to detect a passenger before driving off; but since those are just pressure its easy to throw a bag* or box and suddenly - off goes the car on its own!)



*Which is really annoying when the car is convinced that your backpack is a passenger without its seatbelt on and keeps bleeping at you!

XuQishi wrote:

Traffic jammers will become a viable reality.

Interesting question of ethics, actually. Should an autonomous car always stop for people? If yes, what keeps idiots from causing massive jams by keeping vehicles from moving at all just by dancing around between them?
If people can always just bumble into the streets without looking, city traffic will be ridiculously slow.

Yes an automated car will be easy to deter this way, but I am also talking about abusing networked data from traffic, so thats cars are hacked to drive slower by falsified network information.

Its one of many problems automated vehicles will face and a lot of out of work cabbies and truckers will be happy to help make them a reality.

 Easy E wrote:
Yeah, just ask the Coal Miners!

Sort of. Peak coal mining is the 1970s, unemployment in that era tipped 9%. Right now unemployment is at 4%, and that's after absorbing a lot of women also entering the workforce. So the coal jobs are lost, but other work took its place.

That's been the general pattern, its the basic driver of economic growth - automation and efficiencies mean less people are employed doing task a, the freed up labour goes and works in new task z, and we add the products from task z to overall production and slowly total production grows.

The question is whether the taxi drivers and truck drivers will have a task z to move in to. We've been unsure at every stage before now, but there's always been those new industries. But these days, with the new growth on-line companies employing so few people, maybe it won't happen. Dunno.


Automatically Appended Next Post:

 Yodhrin wrote:
All I meant by that was you can't really use the German economy as a basis for analysing loss of employment due to automation, because the specific effects of the Eurozone on that economy - essentially allowing it to run with a focus on advanced manufacturing with almost none of the accompanying downsides - have massively insulated it from a lot of the secondary effects automation is having on other advanced economies.

In addition to the EU putting Germany in a pretty unique place, it's also a reality that you just can't have more than a couple of countries doing what Germany is doing at any one time. Manufacturing benefits from network effects like any economic sector, the best place for a new factory is frequently right next to the last. Most manufacturing industries just aren't viable as purely domestic production.

Automation in Germany hasn't cost jobs because Germany has instead increased total production and exported more of its production. You can't have other countries copy that, we can't all be net exporters of manufacturing goods.



*The US has some purely domestic industries because there's 330m consumers in a 20 trillion economy, but they're basically it.

China could probably manage it too. The reason they don't now is because they're making more money exporting stuff, primarily to America, than they could make servicing their domestic market.

Yodhrin wrote:All the benefits of automation are accruing to rentier capital, ie the people who own the machines and the people who own the software the machines run. That situation can only continue for so long until it undermines the basic structure of the economy - you can use machines to make things as cheaply as you like, doesn't matter if nobody can earn any money to buy them. The only two practicable solutions are either the institution of substantial, specific, and purposefully unavoidable new taxes on the beneficiaries of automation to fund a universal basic income, or a conscious(and potentially legally-enforced) decision to split up the well-paid high-tech jobs created by automation into multiple reasonably-paid part time jobs so that there's no net loss of employment or earning power, but corporations and politicians and media won't even admit there's a problem let along begin discussing solutions, it's maddening.

When I started this thread, I made a bet with one of my coworkers on how long it would take someone to bring up the whole automation progression into basically "humanity sitting on it's ass while robots do all the labor, forcing us to take rich people's money and give it to us while we do nothing to work" and what country/region that viewpoint would come from.

I won $50.

Orlanth wrote:

XuQishi wrote:

Traffic jammers will become a viable reality.

Interesting question of ethics, actually. Should an autonomous car always stop for people? If yes, what keeps idiots from causing massive jams by keeping vehicles from moving at all just by dancing around between them?
If people can always just bumble into the streets without looking, city traffic will be ridiculously slow.

Yes an automated car will be easy to deter this way, but I am also talking about abusing networked data from traffic, so that cars are hacked to drive slower by falsified network information.

Its one of many problems automated vehicles will face and a lot of out of work cabbies and truckers will be happy to help make them a reality.

And THIS is the biggest problem with AI driven vehicles. People are evil enough now that high school kids have no qualms writing a program that slags peoples' devices just to entertain themselves, and we'll have potentially networked vehicles so that these psychos can rank a confirmed kill? Even looking at networked devices from a different angle, there's still issues. Updates slowing processor speed on a device that depends on its processor speed to make quick judgments in traffic. That can be mitigated with a wipe and reinstall, but that's just going to drive the costs of maintenance even higher.

 Just Tony wrote:
When I started this thread, I made a bet with one of my coworkers on how long it would take someone to bring up the whole automation progression into basically "humanity sitting on it's ass while robots do all the labor, forcing us to take rich people's money and give it to us while we do nothing to work" and what country/region that viewpoint would come from.

I won $50.

Are you seriously disputing that this is the inevitable outcome, given sufficient time?

 Vulcan wrote:
China could probably manage it too. The reason they don't now is because they're making more money exporting stuff, primarily to America, than they could make servicing their domestic market.

CHina is working on expanding its domestic market, but it wouldn't be done to replace exports, but rather to have export and domestic consumption support each other. China is big, but it will be some before China has as massive a car industry. Not in raw numbers, there China is about on par, 4m sales vs 6m in the US, but the average price in the US is $33,000, in China its about $6,000.


Automatically Appended Next Post:

 Just Tony wrote:
When I started this thread, I made a bet with one of my coworkers on how long it would take someone to bring up the whole automation progression into basically "humanity sitting on it's ass while robots do all the labor, forcing us to take rich people's money and give it to us while we do nothing to work" and what country/region that viewpoint would come from.

I won $50.

We are just monkeys to dance for your amusement.

Peregrine wrote:

 Just Tony wrote:
When I started this thread, I made a bet with one of my coworkers on how long it would take someone to bring up the whole automation progression into basically "humanity sitting on it's ass while robots do all the labor, forcing us to take rich people's money and give it to us while we do nothing to work" and what country/region that viewpoint would come from.

I won $50.

Are you seriously disputing that this is the inevitable outcome, given sufficient time?

I think we'll see a global conflict that culls 1/3 or more of the world population before we see fan wielding robots feed us bon bons while we make the 1% run on hamster wheels.

sebster wrote:Automatically Appended Next Post:

 Just Tony wrote:
When I started this thread, I made a bet with one of my coworkers on how long it would take someone to bring up the whole automation progression into basically "humanity sitting on it's ass while robots do all the labor, forcing us to take rich people's money and give it to us while we do nothing to work" and what country/region that viewpoint would come from.

I won $50.

We are just monkeys to dance for your amusement.

I was thinking more along the lines of kinkajou, but the premise is sound...

Which is really annoying when the car is convinced that your backpack is a passenger without its seatbelt on and keeps bleeping at you!

Even more annoying is when it starts doing that without a backpack on the seat. I once had to drive 150 kilometers with the car bleeping at me always getting louder because there was something wrong with the sensor mat in the seat. Of course you can't "recognize" it so it shuts up. Guess how much that improved the safety of my driving that day. I wasn't enraged at all.

When I started this thread, I made a bet with one of my coworkers on how long it would take someone to bring up the whole automation progression into basically "humanity sitting on it's ass while robots do all the labor

I think that is the final goal of automation, though, and I'm not someone to be accused of being a socialist. That would need more of a Star-Trek-like tech level though.

 Just Tony wrote:
I think we'll see a global conflict that culls 1/3 or more of the world population before we see fan wielding robots feed us bon bons while we make the 1% run on hamster wheels.

That's not what I asked. Whether or not a massive war happens has nothing to do with my actual question: do you dispute the inevitability of automation rendering the vast majority of people unemployable in any meaningful fashion, resulting in a welfare state where people are paid for not working?

XuQishi wrote:
Even more annoying is when it starts doing that without a backpack on the seat. I once had to drive 150 kilometers with the car bleeping at me always getting louder because there was something wrong with the sensor mat in the seat. Of course you can't "recognize" it so it shuts up. Guess how much that improved the safety of my driving that day. I wasn't enraged at all.

Why couldn't you just put the seatbelt in to make the car stop beeping?

XuQishi wrote:
There'll be new ones. If mankind is good at anything it's inventing bs jobs so that technical advances never actually reduce the amount of time that people spend at work. We all know that ideally Mr. Robot would do everything so we had leisure time all day, but that's never going to happen. So, taxi drivers, prepare to become Managers of Phone Desinfection (although, as the HHGTTG told us: those guys are really important).

Funny. It already has. Or at least here standard isn't any more 7 days a week work.

 Just Tony wrote:

Orlanth wrote:

XuQishi wrote:

Traffic jammers will become a viable reality.

Interesting question of ethics, actually. Should an autonomous car always stop for people? If yes, what keeps idiots from causing massive jams by keeping vehicles from moving at all just by dancing around between them?
If people can always just bumble into the streets without looking, city traffic will be ridiculously slow.

Yes an automated car will be easy to deter this way, but I am also talking about abusing networked data from traffic, so that cars are hacked to drive slower by falsified network information.

Its one of many problems automated vehicles will face and a lot of out of work cabbies and truckers will be happy to help make them a reality.

And THIS is the biggest problem with AI driven vehicles. People are evil enough now that high school kids have no qualms writing a program that slags peoples' devices just to entertain themselves, and we'll have potentially networked vehicles so that these psychos can rank a confirmed kill? Even looking at networked devices from a different angle, there's still issues. Updates slowing processor speed on a device that depends on its processor speed to make quick judgments in traffic. That can be mitigated with a wipe and reinstall, but that's just going to drive the costs of maintenance even higher.

I think you are getting it wrong here. Yes psychos would love to rack up road accidents for lulz, however I think we can safely assume that vehicle AI design will include safeguards and the safeguards will force the vehicles to be safe. In some ways this is a good thing, however networked data can be manipulated to force the vehicles to be ultra safe to the point of immobility either individually or more likely collectively through traffic.

There is no way around that for programmers as it takes the desired state to an extreme, though yes you could have cases where vehicle AI is hacked for opposite effect, but that will be more difficult.

Even so fatal accidents will occur, when AI feths up, and it does, it goes all the way, my main problem there is there is likely to be a lack of accountability of the operators, or at the very least I don't trust there to be any accountability. Amazon and Google are so big they already don't bother paying tax, once they start running transport they will be professionally disinterested in culpability.

Video of the fatal collision has been released.

I didn't want to watch it, because it seems macabre, but I thought it important to see the evidence of the circumstances of the collision for myself.

Having seen it, I think that part of the blame must lie with the victim. She was pushing her bike across a dark, 40 mph street, with no lights or hi-vis clothing.

I'm not surprised the car did not stop by itself if it relies on radar. A bike is a pretty thin, small object on radar, I imagine. I am a rather more surprised that the safety driver did not see the cyclist and stop the car. However, she seems not to have been paying attention to the road.

The road was dark, the headlights seemed to be in low beam, and the cyclist was wearing dark clothing.

Yup, judging from that video I don't think a human would have been able to stop if they'd been driving, either. The headlights should have been turned up on that road. You couldn't see the person or bike at all before they entered the illumination of the headlights and by that point, at 40mph, there was no time to stop.

So, the headlights should have been turned up and please, people, use lights on your bike and wear high visibility clothing.

Peregrine wrote:

 Just Tony wrote:
I think we'll see a global conflict that culls 1/3 or more of the world population before we see fan wielding robots feed us bon bons while we make the 1% run on hamster wheels.

That's not what I asked. Whether or not a massive war happens has nothing to do with my actual question: do you dispute the inevitability of automation rendering the vast majority of people unemployable in any meaningful fashion, resulting in a welfare state where people are paid for not working?

No, I do not. We don't have ANY reliable AIs in our simple machines that don't need babysat for its daily operations. If we can't get it right on a micro level, thinking it's going to succeed at the mega level is dreaming. I don't deal in dreams, unfortunately. I deal with AI all day at work. Five of the 13 operations I'm signed off on in the machine shop I work in are either robot assisted or a robot cell entirely. Even with the massive resources my corporation has, there are still issues with the robots where they do something absolutely baffling We still have to babysit them. Now, it might be possible that you could have one person babysit the last four operations on my currently assigned line leading up to audit, but none of the other operations can be spread out that thin. AI tech and automation would have to move LEAPS AND BOUNDS to come close to doing 50% of what happens in my shop, let alone 100%.

Why couldn't you just put the seatbelt in to make the car stop beeping?

I did, but it didn't take, there was something wrong with the innards of the seat, I got a new sensor mat and a new belt lock. The most annoying part about that was that nobody ever sits in that seat - it's a coupe, only a small child could even sit there - and I couldn't click the warning away. I wished hard for a "you're wrong, shut up, car"-button.

I'm not surprised the car did not stop by itself if it relies on radar. A bike is a pretty thin, small object on radar, I imagine.

I would think that a bike with a person would have quite the signature. The question is more why that car wasn't also equipped with infra-red as well. I mean, the big advantage of a robot compared to a human is that you can easily make it see in different spectrums at once.

 Kilkrazy wrote:
Video of the fatal collision has been released.

I didn't want to watch it, because it seems macabre, but I thought it important to see the evidence of the circumstances of the collision for myself.

Having seen it, I think that part of the blame must lie with the victim. She was pushing her bike across a dark, 40 mph street, with no lights or hi-vis clothing.

I'm not surprised the car did not stop by itself if it relies on radar. A bike is a pretty thin, small object on radar, I imagine. I am a rather more surprised that the safety driver did not see the cyclist and stop the car. However, she seems not to have been paying attention to the road.

The road was dark, the headlights seemed to be in low beam, and the cyclist was wearing dark clothing.

Wow. She was outright crossing the road in the blind spot between street lamps.

Reminds me of a time I almost hit a deer. You'd think the bastard wouldn't have time to get halfway across the road in the time between the car in front of me passing him but there he was. Except this is obviously more tragic. Not sure I really approve of the safety driver being so inattentive. I don't think she could have prevented the accident cause that woman totally showed up *snaps finger* like that but is constantly looking down and not watching where the car is going really her job?

Driverless cars will always kill people no matter how good the technology become. The low hurdle it must overcome is killing less people than human drivers. That's not going to be difficult when you consider the number of glitches humans have such as drunk driving, txting while driving, falling asleep while driving, road rage, and street racing. Driving will easily be safer once our robotic overlords control our vehicles.

Agrred.

There are two questions in this case, though.

1. Could autonomic cars be equipped with appropriate sensors that would enable them to 'see' a cyclist in the circumstances, and thus avoid a collision?
This is important because we should make the cars as safe as possible. We also need to think about the limits of safety and what demands that may put on other road users, and indeed the limits of usefulness of self-driving cars.

2. What is the point of a "safety driver" if she fails to pay attention?
This is important because the case in favour of testing experimental self-driving cars is based partly on the point that the safety driver will help prevent accidents the cars may cause by their inadequate design.

 Just Tony wrote:
No, I do not. We don't have ANY reliable AIs in our simple machines that don't need babysat for its daily operations. If we can't get it right on a micro level, thinking it's going to succeed at the mega level is dreaming. I don't deal in dreams, unfortunately. I deal with AI all day at work. Five of the 13 operations I'm signed off on in the machine shop I work in are either robot assisted or a robot cell entirely. Even with the massive resources my corporation has, there are still issues with the robots where they do something absolutely baffling We still have to babysit them. Now, it might be possible that you could have one person babysit the last four operations on my currently assigned line leading up to audit, but none of the other operations can be spread out that thin. AI tech and automation would have to move LEAPS AND BOUNDS to come close to doing 50% of what happens in my shop, let alone 100%.

You're making the mistake of assuming that because something isn't great now it must be because of inherent limits, rather than engineering challenges that have not yet been overcome. There is nothing about machine shop work that is inherently impossible for an automated system to do, you're just building objects according to precisely defined specifications. That's exactly the sort of work that machines are great at, as demonstrated by your own admission that you are just babysitting the machines as they do all of the work. The only obstacle to full automation is improving reliability, and the moment the annual scrap losses due to robot mistakes drop below your annual salary you will find yourself without a job. Automated systems are only going to get better and cheaper, so it is only a question of when you become obsolete, not if.

For a counter-example to yours I'll describe the situation at my company (semiconductor manufacturing). I'm in R&D, but I see some of the production side and the majority of their employees are temps whose job consists of loading materials into one side of a machine and unloading them from the other side. For anything more complicated than pressing the "start" button they are under explicit orders to stop and call a tech to do the work for them. Do you honestly think those jobs would survive the introduction of a materials handling robot that costs less than paying a human worker? Of course not. The day that robot exists is the day my company shuts down most of its manufacturing jobs, leaving a handful of technicians supervising an automated factory and a handful of engineers and managers handling the design and organizational work.

So what happens when this kind of thing happens across the entire economy, and unskilled labor is replaced with machines? Most of these people aren't going to get the kind of engineering/management/etc jobs that will require humans (at least for the foreseeable future) because you're largely talking about people with a high school diploma at most, or maybe a university degree of the "well, you officially spent four years at a university" sort that doesn't provide any real job skills. I know plenty of people who have zero hope of handling the kind of math and science classes that I had to take to get my engineering degree, the kind of degree that you'll need to qualify for those surviving jobs. So what are they supposed to do, starve to death because they weren't good at calculus?

Computers are very good at calculus.

The thing that will distinguish future human work from automation is the ability to frame problems to discover the best solution from the human perspective.

For example, right now an architect will make a design to best meet his client's needs, then feed the specs into a spreadsheet to work out the cost of buying all the materials.

Also "creativity", and the ability to think flexibly, and deal with new situations on the fly.

 Kilkrazy wrote:
Computers are very good at calculus.

Yes, but you still need a human who understands calculus to run the computer. Without that knowledge the computer is worthless, as you don't know why it is doing what it does and can't get meaningful engineering solutions from it. You aren't hiring people with science/engineering degrees to do math problems, you're hiring them for the skills they learned after taking 2-3 years of calculus classes to understand the real work of their field.

 Just Tony wrote:

I think we'll see a global conflict that culls 1/3 or more of the world population before we see fan wielding robots feed us bon bons while we make the 1% run on hamster wheels.

The rich would indeed rather kill a third of the planet than share wealth.

Indeed. And unless you are dealing with quantum effects or other stuff like that (such as general relativity field equations etc.) most of the problems you face are unlikely to be more complex than a first or second order differential equation, which are pretty easy to solve (especially second order which are, funnily, easier than first order as the method for solving them is always the same), even for humans. For others it is often a case of manipulating them until they are in a solvable form, at which point ypu just look up the answer in a book/the internet.

That rearranging requires imagination, intuition and an understanding of the science/maths behind the problem.

As usual,the BBC has a timely article on this topic.

http://www.bbc.co.uk/news/business-43259903

Wanted: Robot wrangler. No experience required.

 schadenfreude wrote:
Driverless cars will always kill people no matter how good the technology become. The low hurdle it must overcome is killing less people than human drivers. That's not going to be difficult when you consider the number of glitches humans have such as drunk driving, txting while driving, falling asleep while driving, road rage, and street racing. Driving will easily be safer once our robotic overlords control our vehicles.

There is a caveat to that. Only if the vast majority/all of the other cars on the road are also driverless cars and all of them communicate with each other.

A mixture of driverless cars and regular cars will likely result in massive glitches on the part of the driverless cars as they will not interact well with human drivers, while human drivers on the other hand are better at reacting to unexpected things. A driverless car can only act within the limits of its programming. So if something unexpected happens, like say a sudden lane change from another vehicle, a human might be able to compensate while a driverless car might not compensate enough. If its programming makes it brake too much or too little it could cause a chain reaction.

The transition period from where we are now to 100% automated cars is where the carnage will happen. And the problem will simply rest on upsetting the current status quo, IE: the blame will be on driverless cars.

In some ways, its like transitioning to Communism. The destination seems idyllic and in many ways it is a perfect society in theory. But the journey there is simply too full of pitfalls and carnage to make the journey worthwhile.