If I take my hands off the wheel, my starts or veer slightly to the right. I thought about getting it fixed (and at some point it may be fixed when doing a wheel alignment) but chose not to since it makes me keep my hands on the wheel and actually looking where I am going.
This is in a car that does not have any automatic features (other than automatic choke in the carburetor).
Well, one might think that "keeping constant speed" and "not going into a ditch" are separate functions of the car, especially if they are turned on separately. So, you might assume that it is possible to turn off one but keep the other.
What it should do is give control back to the driver, but only notify about this by turning on a lamp on the dashboard or projecting something on the windshield. If the driver fails to steer the car to keep it in the lane (if the road is very straight, the car could start to veer very lightly to the right making the driver correct for it), the car should stop. This should happen at random intervals and the driver would have to manually control the car for random amount of time (a few minutes).
In this situation, the driver would have to actually pay attention to the road (and dashboard) instead of waiting for an audible signal or randomly wiggling the steering wheel.
AFAIK the long highways in the US are very straight compared to various inter-city roads in Europe, the curving roads also force the driver to pay more attention to the road and gives him something to do (correct the trajectory) so he is less tempted to let go of the wheel and browse the net..
I mean it is almost like having a self-driving car that you do not need to pay attention to the road most of the time and the car will not go into a ditch or rear-end another car.
So you can get bored by having absolutely nothing to do, so the temptation of doing something is just too great. And it works most of the time, the autopilot keeps the car on the road and avoids danger.
Except for that 0.01% when it fails and you have to react as quickly as if you have been driving all this time.
When driving regular car, you have to make frequent minor adjustments to keep the car on the road (the road isn't straight after all), so there is less time to get as bored as when you have nothing to do.
My grandfather worked as a bus driver for a while (driving between cities), he told me that the road from Kaunas to Vilnius (in Lithuania) was too straight for him and he had trouble not falling asleep at the wheel (so he used to talk to the passengers etc and never actually fell asleep), while driving to other cities was easier because the road is not as straight.
Using the autopilot most likely looks like driving on a completely straight road with a car that does not veer to any direction by itself.
Yea, it also may be very difficult to program such things (after all, there are things that computers can be easily programmed to do, but humans find it difficult to do themselves (complex calculations for example), and conversely there are things that are quite easy for humans, but very difficult to program a computer to do (speech/object recognition for example)).
Also, there may be a time when it is preferable to avoid or run a pedestrian over as opposed to stopping (if said pedestrian (and maybe some of his friends) is standing in the middle of the road with his gun pointed at the incoming car and is not wearing a uniform of a group that is legally allowed to do that (police, special forces)).
No "truck crossing" signs in my country. But if I see a car approaching an intersection, I slow down just in case the other driver does not stop and let me trough (assuming I have right of way).
Still, a truck on the road is very obvious. Unless it was covered by trees or a building just before entering the path of the car, I would have noticed the truck getting closer and started to slow down ("hey, maybe he is gong to turn even if the signal is red for him").
I always slow down (and place my foot on the brake) when I see a car approach an intersection from the side road just in case the other driver does not notice me.
However, the driver may have noticed the truck if he was actually driving. Since the car was on autopilot, the driver probably was browsing the net or looking in directions other than the road, after all, the car is driving itself, there is nothing else to do.
Either I am driving or someone else is. Just like I do not pay a lot of attention to the road when I am taking the taxi (nor do I apply the handbrake if I see something on the road, I would hope the taxi driver would stop the car in time and me applying the handbrake may make things worse).
What is suggested here, is taking a taxi where the driver is drunk or very sleepy ("if you see m going straight at a wall, apply the handbrake and slap me to wake me up"), so here is the worst of both words - I cannot relax and ignore the road (like I would be if driven by another human) and yet, I am not in control, so I have nothing to do and am really bored and want to browse the net or do something else.
Cruise control only keeps your speed constant, you still have to pay attention to the road or will drive into a ditch when the road turns.
I use cruise control because keeping my leg locked in a single position for long periods is physically uncomfortable.
My car has a spring loaded accelerator pedal, where (maybe by coincidence, maybe by design), the spring can support the wight of my foot at any position of the throttle, I can essentially rest my foot there.
Driving a different car I noticed that it is different, here, if I rest my foot on the pedal, it goes full open.
I am not saying that cruise control is bad, but I have not noticed a need for it, at least driving up to three hours at a time. Maybe taking a longer trip on a road where the speed does not change it may become uncomfortable, I do not know.
However, once you turn cruise control on, where do you place your foot? If you needed to suddenly brake (moose on the road etc), will it take more or less time to slam the brake compared to if you kept your foot on the accelerator?
If course, if the driver is busy writing a text message or browsing the net (or drunk, or asleep) then he may not even notice a cargo ship.
However, when this happens, the driver is (correctly) blamed and ridiculed for it, since normally you may not notice a cyclist or a pedestrian (dressed in black clothes at night with no reflectors), but a tractor trailer is kinda obvious.
This is why it is bad if the computer makes a mistake most human drivers wouldn't. I am sure it normally has much better visibility and reaction time than most humans, but still, the autopilot did not notice a lorry.
If the accident was caused by the computer in circumstances where most human drivers would also cause an accident (like failing to notice a ninja cosplayer (in my country we call such pedestrians "bowling pins") on an unlit road), it would not be much of a story, just "meh, I would not have done better in that case".
The problem is that the computer crashed the car in a very stupid way that most likely would have been prevented if a human was driving the car. I mean the computer failed to notice a huge lorry. I would have noticed it even without my glasses.
If the accident was something more like the ones sober people cause, it would not be such big news. However, if a human driver caused such an accident, people would think that he was really inattentive (searching for his phone on the floor, texting etc) or drunk.
So, a self-driving car caused an accident that only a drunk/texting driver would cause.
Even if the driver was paying attention (in that case what is the point of the autopilot?), how is he supposed to know that this is the one time the computer is going to fail to stop the car? I mean similar situations, but with other cars or pedestrians may have happened before and the computer stopped the car. So, one may believe that if the computer can see a pedestrian, then it most definitely can see a lorry which is may times bigger than a pedestrian. By the time you notice that the computer is not going to stop the car, it's too late.
Having the car on autopilot, but requiring the driver to "pay attention" and be ready to take over within seconds is the worst combination possible.
If the car did not have autopilot, the driver would have been more attentive, since he would have been driving the car. If the car was completely on autopilot, then one hopes the computer would detect objects on the road, at least the bigger ones, like lorries and tanks.
What is the reason one would switch the car to autopilot? Most likely so that they can be less attentive to the road. If I need to be as attentive as driving an older car, then I will not use the autopilot. The reason is that the constant minor adjustments I usually have to make (the road is not straight after all) help me to keep my attention on the road.
If autopilot means that the driver has nothing to do most of the time, but has to react really fast when something bad happens is a problem, because boredom reduces the attentiveness.
Also, let's say the driver was paying attention to the road. The car is on autopilot, he sees that a lorry is getting closer to him. How is he supposed to know that this is the time the computer will fail to notice a huge lorry (it noticed much smaller cars and pedestrians with no problems before) and take over?
This is assuming I am not liable for it (since someone else is controlling the car at the time). If I am liable, then it depends on the punishment for breaking someone's leg (if they are at fault then screw them), I would rather destroy my car than go to jail for example.
Which is why I am waiting for a terrorist attack by hacking insecure cars and causing them to go in a ditch or run over pedestrians. If such a thing could be arranged on multiple cars at once worldwide (and it seems that one day it will be possible, seeing how car manufacturers fail to secure their cars and how cars become increasingly computer-driven and connected to the internet), it would cause mass panic, since people would become afraid of their own cars.
Also, we should stop letting everybody into the country without extensive background checks. Also, we should consider banning anyone from certain countries (that are unstable, in the middle of a civil war or have very high rates of terrorism, you know, countries like Syria or Libya) from entering.
I am all for socialism and equality, but only if everyone is truly equal and there are no freeloaders (or 1%, sharing means sharing for everyone, not just the working class that cannot evade taxes). Welfare system is needed as a safety net, but it should primarily be used for helping people with disabilities, sick people etc. Everybody else should have to work - unless they save up enough money.
The USSR had a few good ideas IMO (and lots of bad ones). One of them was that everybody had to work - you were provided with a job by the government (and could change jobs if you wanted), if you went to university, you got a job that was based on your degree. You could choose to not work, but you did not receive welfare (assuming you are healthy and younger than the retirement age) and getting caught sleeping in a park meant getting a job (the government was happy to provide you with one).
As for other countries, the US should consider occupying a country, instead of blowing it up and leaving. While democracy is a good thing, people, who have been living under dictatorship for generations do not really have a concept for it (just look at Russia - they never had real democracy and for a lot of them, a good leader means one with a big iron fist). So, if the current situation of a country requires you blowing it up, then you should install a very strict government (loyal to you), create order (after all, there is a lot of chaos after a war) and then allow the people to have increasingly more say in what the country does, maybe letting them elect city councils at first, then some time later parts of parliament etc.
Otherwise, after, you destroy the law enforcement and the government criminal organizations become the most organized and well armed groups in the country (after all, they managed to fight the cops when the previous government was in power) and will take over. See Iraq, Libya and Syria for examples.
Of course outcome #1 is way more preferable than any other and I hope nobody wants to hit someone else if it was possible to completely avoid the accident.
But, bad luck may happen. Brake failure at the most inappropriate time (kinda happened to me - car in front of me suddenly stopped, I tried to stop and one front wheel brake disk broke, fortunately the other brakes worked well enough for me to stop, though it was not fun going back home with only three working brakes).
Some other failure can cause the car not be able to stop in time or actively need to get out of the way (car, driving in the opposite direction suddenly swerved into my lane and is now coming towards me).
you would rather break your leg than cause easily repaired damage to somebody else's car.
Not easily repairable damage, massive damage. Easily repairable damage is #2 on the list.
And, yes, I would expect the other car owner to be selfish (just like I am) and choose to break my leg over destroying his car.
The thing is, I prefer reversible (easier repairable) damage over irreversible. This is why I would prefer a bump on the head that heals completely and without lasting effects instead of a big dent on the car (which will not heal by itself). This is also why I would prefer a more difficult to repair injury (but again, not permanent disability or something like that) instead of a huge damage to the car (needing to replace large parts of the body).
Then I would prefer injuring others instead of myself (or damaging my car a lot).
Obviously, I would rather the car get destroyed instead of killing or permanently disabling someone. But if a death or disability is going to happen either way, I'd prefer if it does not happen to me.
I care about my car more than about strangers. It's not just money for me (as evidenced by me paying about the same amount of money to repair my car after it was rear-ended (I stopped to let a pedestrian cross, the guy behind me didn't notice) as I would have paid to buy a very similar car).
Since the manufacturer of the car is going to be liable for the injury (I am assuming this, just imagining an AI as a human driver I hired to drive me around in my car - if he hits something, he is liable, not me), then I'd rather they paid for the injury than me having to replace my car (especially if the model is not made anymore and I am forced to get a newer one).
If I am liable for the damage, assuming this is known in advance and I have time to think about it (which I probably won't in a real accident with regular cars, but the computer will), I would have to look up what the punishment is for this - I'd rather destroy the car than go to jail, but I also would rather pay 10kEUR to keep my car (even though the same year/model new one costs maybe 2kEUR).
But if somebody else (my hired driver or the car manufacturer) is paying for the accident - I want to keep my car.
Then again, I would have to be forced by law to either get a self-driving car or a car made after 2000, though I would prefer cars made before 1985.
If gasoline powered cars are banned from cities in my country, I would look into converting my current car to electric power first (assuming that was allowed by law), the technology would probably have improved enough to have useful range even with a less aerodynamic body and the electric motor just attached to the driveshaft or transmission.
It is also you who put the car in a situation where you have to choose between hitting the pedestrian and the tree.
The pedestrian could have jumped out in front of me, not on a crosswalk, from the front of a stopped lorry (that is, where it was impossible for me to see him). Or he could even be doing this deliberately to try to get money from my insurance (I know one who tried that, he was found out, so, not only he was severely injured, he had to pay for the damage to the car).
And no, I will not always go for the tree. And if I am not in control of the car, then I prefer that the driver (be it a human or AI) goes for the pedestrian to the point of not buying a self-driving car with a preference for the tree. I paid money for it, it should protect my life.
I did not say that I can carefully consider all the outcomes before deciding whether to hit a tree or a man. The time is usually too short to consider anything, other than trying to stop and maybe turning the car in a direction away from any object (maybe unsuccessfully).
However, computers do what they are told and AI most likely does have time to consider this. Which means that this is now a problem - I do not want the AI in my car (or a taxi driver) to carefully consider the outcomes and decide to kill me instead of a pedestrian. Since it is most likely impossible for the taxi driver to carefully consider all options, I accept that the outcome is gong to be random (he may be too slow to react and hit the object in front whether it's a tree or a man, he may try to avoid the man in front only to hit a tree he didn't notice or he might try to avoid hitting the tree only to hit the man).
Not so when such situations are considered well in advance (when programming the AI) - in that case I will not want to ride in a car that is driven by AI that will predictably choose to hit a tree instead of a man.
For the purposes of the example, assume that the speed is high enough that hitting a tree will kill or permanently disable the people in the car, while hitting the man will kill the man, but leave the passengers better off (without permanent disability).
In addition to that, when I am driving, I am in control and responsible for my decisions (whether they are thought out or I was just too slow to react). Not so, when the AI is in control.
They probably thought they would survive hitting the tree or building. Hitting people means going to jail.
OTOH, it I have no control of my car (it being self-driving and all), then I would prefer these outcomes (in order of preference): 1. No damage to anyone (safe stop) 2. Easily repaired damage to the car. 3. Very small self-healing injuries to me. 4. Non-permanent injuries to other people (broken leg etc). 5. Massive damage to the car. 6. Killing other people. 7. Killing me.
Actually, if I was a passenger in a taxi, I would prefer the taxi driver mows down a lot of pedestrians instead of driving into a wall and killing me.
The difference is that if somebody else does it (either the taxi driver or the AI in the car) ten I am not liable for it.
Also, those young people who really wanted to stay couldn't be bothered to actually go and vote. If they did and the same proportion was for remaining, the remain side would have won.
You have the right to not vote. However, if you exercise that right, you lose the right to bitch about the outcome, because your vote actually was "I don't care, do whatever you want".
Slashdot Mirror
If I take my hands off the wheel, my starts or veer slightly to the right. I thought about getting it fixed (and at some point it may be fixed when doing a wheel alignment) but chose not to since it makes me keep my hands on the wheel and actually looking where I am going.
This is in a car that does not have any automatic features (other than automatic choke in the carburetor).
Well, one might think that "keeping constant speed" and "not going into a ditch" are separate functions of the car, especially if they are turned on separately. So, you might assume that it is possible to turn off one but keep the other.
What it should do is give control back to the driver, but only notify about this by turning on a lamp on the dashboard or projecting something on the windshield. If the driver fails to steer the car to keep it in the lane (if the road is very straight, the car could start to veer very lightly to the right making the driver correct for it), the car should stop. This should happen at random intervals and the driver would have to manually control the car for random amount of time (a few minutes).
In this situation, the driver would have to actually pay attention to the road (and dashboard) instead of waiting for an audible signal or randomly wiggling the steering wheel.
AFAIK the long highways in the US are very straight compared to various inter-city roads in Europe, the curving roads also force the driver to pay more attention to the road and gives him something to do (correct the trajectory) so he is less tempted to let go of the wheel and browse the net..
I mean it is almost like having a self-driving car that you do not need to pay attention to the road most of the time and the car will not go into a ditch or rear-end another car.
So you can get bored by having absolutely nothing to do, so the temptation of doing something is just too great. And it works most of the time, the autopilot keeps the car on the road and avoids danger.
Except for that 0.01% when it fails and you have to react as quickly as if you have been driving all this time.
When driving regular car, you have to make frequent minor adjustments to keep the car on the road (the road isn't straight after all), so there is less time to get as bored as when you have nothing to do.
My grandfather worked as a bus driver for a while (driving between cities), he told me that the road from Kaunas to Vilnius (in Lithuania) was too straight for him and he had trouble not falling asleep at the wheel (so he used to talk to the passengers etc and never actually fell asleep), while driving to other cities was easier because the road is not as straight.
Using the autopilot most likely looks like driving on a completely straight road with a car that does not veer to any direction by itself.
Yea, it also may be very difficult to program such things (after all, there are things that computers can be easily programmed to do, but humans find it difficult to do themselves (complex calculations for example), and conversely there are things that are quite easy for humans, but very difficult to program a computer to do (speech/object recognition for example)).
Also, there may be a time when it is preferable to avoid or run a pedestrian over as opposed to stopping (if said pedestrian (and maybe some of his friends) is standing in the middle of the road with his gun pointed at the incoming car and is not wearing a uniform of a group that is legally allowed to do that (police, special forces)).
No "truck crossing" signs in my country. But if I see a car approaching an intersection, I slow down just in case the other driver does not stop and let me trough (assuming I have right of way).
Still, a truck on the road is very obvious. Unless it was covered by trees or a building just before entering the path of the car, I would have noticed the truck getting closer and started to slow down ("hey, maybe he is gong to turn even if the signal is red for him").
I always slow down (and place my foot on the brake) when I see a car approach an intersection from the side road just in case the other driver does not notice me.
However, the driver may have noticed the truck if he was actually driving. Since the car was on autopilot, the driver probably was browsing the net or looking in directions other than the road, after all, the car is driving itself, there is nothing else to do.
Either I am driving or someone else is. Just like I do not pay a lot of attention to the road when I am taking the taxi (nor do I apply the handbrake if I see something on the road, I would hope the taxi driver would stop the car in time and me applying the handbrake may make things worse).
What is suggested here, is taking a taxi where the driver is drunk or very sleepy ("if you see m going straight at a wall, apply the handbrake and slap me to wake me up"), so here is the worst of both words - I cannot relax and ignore the road (like I would be if driven by another human) and yet, I am not in control, so I have nothing to do and am really bored and want to browse the net or do something else.
Cruise control only keeps your speed constant, you still have to pay attention to the road or will drive into a ditch when the road turns.
I use cruise control because keeping my leg locked in a single position for long periods is physically uncomfortable.
My car has a spring loaded accelerator pedal, where (maybe by coincidence, maybe by design), the spring can support the wight of my foot at any position of the throttle, I can essentially rest my foot there.
Driving a different car I noticed that it is different, here, if I rest my foot on the pedal, it goes full open.
I am not saying that cruise control is bad, but I have not noticed a need for it, at least driving up to three hours at a time. Maybe taking a longer trip on a road where the speed does not change it may become uncomfortable, I do not know.
However, once you turn cruise control on, where do you place your foot? If you needed to suddenly brake (moose on the road etc), will it take more or less time to slam the brake compared to if you kept your foot on the accelerator?
If course, if the driver is busy writing a text message or browsing the net (or drunk, or asleep) then he may not even notice a cargo ship.
However, when this happens, the driver is (correctly) blamed and ridiculed for it, since normally you may not notice a cyclist or a pedestrian (dressed in black clothes at night with no reflectors), but a tractor trailer is kinda obvious.
This is why it is bad if the computer makes a mistake most human drivers wouldn't. I am sure it normally has much better visibility and reaction time than most humans, but still, the autopilot did not notice a lorry.
If the accident was caused by the computer in circumstances where most human drivers would also cause an accident (like failing to notice a ninja cosplayer (in my country we call such pedestrians "bowling pins") on an unlit road), it would not be much of a story, just "meh, I would not have done better in that case".
The problem is that the computer crashed the car in a very stupid way that most likely would have been prevented if a human was driving the car. I mean the computer failed to notice a huge lorry. I would have noticed it even without my glasses.
If the accident was something more like the ones sober people cause, it would not be such big news. However, if a human driver caused such an accident, people would think that he was really inattentive (searching for his phone on the floor, texting etc) or drunk.
So, a self-driving car caused an accident that only a drunk/texting driver would cause.
Even if the driver was paying attention (in that case what is the point of the autopilot?), how is he supposed to know that this is the one time the computer is going to fail to stop the car? I mean similar situations, but with other cars or pedestrians may have happened before and the computer stopped the car. So, one may believe that if the computer can see a pedestrian, then it most definitely can see a lorry which is may times bigger than a pedestrian. By the time you notice that the computer is not going to stop the car, it's too late.
Having the car on autopilot, but requiring the driver to "pay attention" and be ready to take over within seconds is the worst combination possible.
If the car did not have autopilot, the driver would have been more attentive, since he would have been driving the car.
If the car was completely on autopilot, then one hopes the computer would detect objects on the road, at least the bigger ones, like lorries and tanks.
What is the reason one would switch the car to autopilot? Most likely so that they can be less attentive to the road. If I need to be as attentive as driving an older car, then I will not use the autopilot. The reason is that the constant minor adjustments I usually have to make (the road is not straight after all) help me to keep my attention on the road.
If autopilot means that the driver has nothing to do most of the time, but has to react really fast when something bad happens is a problem, because boredom reduces the attentiveness.
Also, let's say the driver was paying attention to the road. The car is on autopilot, he sees that a lorry is getting closer to him. How is he supposed to know that this is the time the computer will fail to notice a huge lorry (it noticed much smaller cars and pedestrians with no problems before) and take over?
This is assuming I am not liable for it (since someone else is controlling the car at the time). If I am liable, then it depends on the punishment for breaking someone's leg (if they are at fault then screw them), I would rather destroy my car than go to jail for example.
Which is why I am waiting for a terrorist attack by hacking insecure cars and causing them to go in a ditch or run over pedestrians. If such a thing could be arranged on multiple cars at once worldwide (and it seems that one day it will be possible, seeing how car manufacturers fail to secure their cars and how cars become increasingly computer-driven and connected to the internet), it would cause mass panic, since people would become afraid of their own cars.
Also, we should stop letting everybody into the country without extensive background checks. Also, we should consider banning anyone from certain countries (that are unstable, in the middle of a civil war or have very high rates of terrorism, you know, countries like Syria or Libya) from entering.
I am all for socialism and equality, but only if everyone is truly equal and there are no freeloaders (or 1%, sharing means sharing for everyone, not just the working class that cannot evade taxes). Welfare system is needed as a safety net, but it should primarily be used for helping people with disabilities, sick people etc. Everybody else should have to work - unless they save up enough money.
The USSR had a few good ideas IMO (and lots of bad ones). One of them was that everybody had to work - you were provided with a job by the government (and could change jobs if you wanted), if you went to university, you got a job that was based on your degree. You could choose to not work, but you did not receive welfare (assuming you are healthy and younger than the retirement age) and getting caught sleeping in a park meant getting a job (the government was happy to provide you with one).
As for other countries, the US should consider occupying a country, instead of blowing it up and leaving. While democracy is a good thing, people, who have been living under dictatorship for generations do not really have a concept for it (just look at Russia - they never had real democracy and for a lot of them, a good leader means one with a big iron fist). So, if the current situation of a country requires you blowing it up, then you should install a very strict government (loyal to you), create order (after all, there is a lot of chaos after a war) and then allow the people to have increasingly more say in what the country does, maybe letting them elect city councils at first, then some time later parts of parliament etc.
Otherwise, after, you destroy the law enforcement and the government criminal organizations become the most organized and well armed groups in the country (after all, they managed to fight the cops when the previous government was in power) and will take over. See Iraq, Libya and Syria for examples.
Of course outcome #1 is way more preferable than any other and I hope nobody wants to hit someone else if it was possible to completely avoid the accident.
But, bad luck may happen. Brake failure at the most inappropriate time (kinda happened to me - car in front of me suddenly stopped, I tried to stop and one front wheel brake disk broke, fortunately the other brakes worked well enough for me to stop, though it was not fun going back home with only three working brakes).
Some other failure can cause the car not be able to stop in time or actively need to get out of the way (car, driving in the opposite direction suddenly swerved into my lane and is now coming towards me).
So, all outcomes happen sometimes.
you would rather break your leg than cause easily repaired damage to somebody else's car.
Not easily repairable damage, massive damage. Easily repairable damage is #2 on the list.
And, yes, I would expect the other car owner to be selfish (just like I am) and choose to break my leg over destroying his car.
The thing is, I prefer reversible (easier repairable) damage over irreversible. This is why I would prefer a bump on the head that heals completely and without lasting effects instead of a big dent on the car (which will not heal by itself). This is also why I would prefer a more difficult to repair injury (but again, not permanent disability or something like that) instead of a huge damage to the car (needing to replace large parts of the body).
Then I would prefer injuring others instead of myself (or damaging my car a lot).
Obviously, I would rather the car get destroyed instead of killing or permanently disabling someone. But if a death or disability is going to happen either way, I'd prefer if it does not happen to me.
I care about my car more than about strangers. It's not just money for me (as evidenced by me paying about the same amount of money to repair my car after it was rear-ended (I stopped to let a pedestrian cross, the guy behind me didn't notice) as I would have paid to buy a very similar car).
Since the manufacturer of the car is going to be liable for the injury (I am assuming this, just imagining an AI as a human driver I hired to drive me around in my car - if he hits something, he is liable, not me), then I'd rather they paid for the injury than me having to replace my car (especially if the model is not made anymore and I am forced to get a newer one).
If I am liable for the damage, assuming this is known in advance and I have time to think about it (which I probably won't in a real accident with regular cars, but the computer will), I would have to look up what the punishment is for this - I'd rather destroy the car than go to jail, but I also would rather pay 10kEUR to keep my car (even though the same year/model new one costs maybe 2kEUR).
But if somebody else (my hired driver or the car manufacturer) is paying for the accident - I want to keep my car.
Then again, I would have to be forced by law to either get a self-driving car or a car made after 2000, though I would prefer cars made before 1985.
If gasoline powered cars are banned from cities in my country, I would look into converting my current car to electric power first (assuming that was allowed by law), the technology would probably have improved enough to have useful range even with a less aerodynamic body and the electric motor just attached to the driveshaft or transmission.
It is also you who put the car in a situation where you have to choose between hitting the pedestrian and the tree.
The pedestrian could have jumped out in front of me, not on a crosswalk, from the front of a stopped lorry (that is, where it was impossible for me to see him). Or he could even be doing this deliberately to try to get money from my insurance (I know one who tried that, he was found out, so, not only he was severely injured, he had to pay for the damage to the car).
And no, I will not always go for the tree. And if I am not in control of the car, then I prefer that the driver (be it a human or AI) goes for the pedestrian to the point of not buying a self-driving car with a preference for the tree. I paid money for it, it should protect my life.
I did not say that I can carefully consider all the outcomes before deciding whether to hit a tree or a man. The time is usually too short to consider anything, other than trying to stop and maybe turning the car in a direction away from any object (maybe unsuccessfully).
However, computers do what they are told and AI most likely does have time to consider this. Which means that this is now a problem - I do not want the AI in my car (or a taxi driver) to carefully consider the outcomes and decide to kill me instead of a pedestrian. Since it is most likely impossible for the taxi driver to carefully consider all options, I accept that the outcome is gong to be random (he may be too slow to react and hit the object in front whether it's a tree or a man, he may try to avoid the man in front only to hit a tree he didn't notice or he might try to avoid hitting the tree only to hit the man).
Not so when such situations are considered well in advance (when programming the AI) - in that case I will not want to ride in a car that is driven by AI that will predictably choose to hit a tree instead of a man.
For the purposes of the example, assume that the speed is high enough that hitting a tree will kill or permanently disable the people in the car, while hitting the man will kill the man, but leave the passengers better off (without permanent disability).
In addition to that, when I am driving, I am in control and responsible for my decisions (whether they are thought out or I was just too slow to react). Not so, when the AI is in control.
They probably thought they would survive hitting the tree or building. Hitting people means going to jail.
OTOH, it I have no control of my car (it being self-driving and all), then I would prefer these outcomes (in order of preference):
1. No damage to anyone (safe stop)
2. Easily repaired damage to the car.
3. Very small self-healing injuries to me.
4. Non-permanent injuries to other people (broken leg etc).
5. Massive damage to the car.
6. Killing other people.
7. Killing me.
Actually, if I was a passenger in a taxi, I would prefer the taxi driver mows down a lot of pedestrians instead of driving into a wall and killing me.
The difference is that if somebody else does it (either the taxi driver or the AI in the car) ten I am not liable for it.
Also, those young people who really wanted to stay couldn't be bothered to actually go and vote. If they did and the same proportion was for remaining, the remain side would have won.
You have the right to not vote. However, if you exercise that right, you lose the right to bitch about the outcome, because your vote actually was "I don't care, do whatever you want".