Skip to main content

Week 4 [16.04-22.04.2018] Autonomous cars are among us. But are their decisions autonomous?


Week 4 [16.04-22.04.2018] Autonomous cars are among us. But are their decisions autonomous?

Some time ago autonomous cars were something that was even hard to imagine. Nowadays autonomous cars are real and each of us can have one (of course if you have enough $$$) These cars have a lot of CPUs, sensors, different algorithms etc. As we can see it’s possible to program them to deal with casual situations, even difficult ones. But there seems to be tough edge case to consider. Each of us sometimes has to choose the lesser of two evils. Imagine that you are driving and another car in front of you brakes hard.  It’s obvious that you will try to avoid hitting him but on your left side you have a car with a passenger without seatbelts fastened and on your right side you see a motorcyclist without a helmet. What to do?! There are a lot of social dilemmas like this. Should your own car do everything to protect you or maybe it should also think about others?! Watch this short video and share your opinion with us!


Comments

Unknown said…
I agree with the speaker, that problem behind regulation of autonomous cars is not something that can be easily solved in a technical way using complex algorithms . Situation like that shown in the video requires from car's AI to make a moral choice, which machine itself cannot make without any human intervention or guidance. It is our responsibility to make such decisions. But as it was shown people's opinion can also vary. I think that for now we should continue doing researches involving artificial intelligence and continue improving cars' safety. Because of the complexity of this dilemma that easiest way to avoid this event is to minimize a probability of occurring it in real life.
Lukasz Mroczek said…
Good to read that you agree! This area of autonomous cars definitely needs much more research and it's worth it!
Unknown said…
I think it's way to much thinking. Going this road I should start analyzing if don't harm humanity buying certain milk or breathing too fast.

I believe that no robot should have the ability to harm human, especially not for the 'common good'. Not likely, bu still, tech can always be exploited. Somehow, I have a feeling that the speaker would like to live in a place when nobody dies or is killed, which is pretty abstract and in my opinion delusive. I hope he doesn't.

I agree we need to research and improve. As for myself, I wouldn't drive such car because I simply enjoy driving. If I go too far to drive I make stops, fly by plane or go by train. One good point to use his technology is it could minimize costs and losses of companies in truck business, but I've heard lots of opinions from people who know something about this work and they always list countless cases where today's tech wouldn't do well.
Lukasz Mroczek said…
Maybe it's a lot of thinking but it's very important. We have to take it into consideration while programming and developing autonomous cars. Maybe some time in near future there will be solution for problems like this?
Unknown said…
It’s complex dilemma. In presented situation it may seem that the person that was in the car should have taken responsibility for the accident that vehicle has caused. Since they were the once that chose to use the car. However, if in the car were more people than from societal perspective it would seem reasonable to choose the one bystander instead. Apparently, people in the survey from presentation overwhelmingly chose to be the ones the are protected in given scenario. I think that in order not to halt the adaptation of autonomous cars it would be better to instruct the machine to prioritize passengers, at least at the beginning. Just to sooth their concerns. It would result in saving more lives as autonomous car are bound to have less accidents. And the problem presented in video would be even less likely to even happen.
Lukasz Mroczek said…
Of course I agree that it's complex dilemma. Nevertheless it's important and designers, developers etc have to take it into consideration.
Filip Sawicki said…
Nobody really knows how to regulate autonomous vehicles and, as in presented situation, who truly is responsible for incidents with participation of artificial intelligence. This problem will only grow in the future as more and more devices will incorporate reasoning processes to some degree. The only think we can do is to audit the development and implementation of AI with a hope of minimizing the harm it will unintentionally create. In my opinion in situations of inevitable accident the algorithm should calculate probability for each foreseeable event and choose the one which will lead to the smallest number of casualties.
Jakub Nietupski said…
I don't think that there is a dilemma at all - if I were to buy an autonomous car I would definitely want it to protect my live under any circumstances. Because of that, the engineers will design cars in such a way that satisfies the demand. There is another problem though - whether we should regulate car manufactures by law so that autonomous cars choose to minimize harm over protecting their owner. But I think it's not easy to establish what the minimizing harm really is. The car during an accident will not have objective, complete information about the whole situation so if it always chooses to protect itself then it's behavior then it's predictable for other vehicles and thus can contribute to lowering the death toll.
Anna Koca said…
Living in the information era, we will have to face the question of AI morality more and more often, so we might as well start to think on this subject. It will begin with self-driven cars, and end on robot nannies or brain tumor operations supported by AI solutions. It is a question of trust, and I think that it may take (maybe even must) several errors and possibly victims, to improve artificial intelligence to the level where we can truly put ourselves in its hands. But the real question is: isn't it worth it? Human doctors make mistake - they get tired, their hands may experience an unexpected cramp - but we do not judge them nearly as harsh. In the long run, I guess that AI solutions are far less dangerous and we should aim to take it to the level of perfection.
Lukasz Mroczek said…
I agree with you but there is one more thing. Algorithm will calculate probability and choose the best way but unfortunately this algorithm can't take emotions into consideration so there will be always someone who will be disappointed.
Lukasz Mroczek said…
Kinda egoistic approach :) But I do agree that it'd great idea to introduce "standard" that every autonomous car will behave in nearly the same way in particular situations.
Lukasz Mroczek said…
I think you're right. Some people have negative approach to AI but in some cases it's much better and also more save than old solutions. It's definitely worth developing and improving.
Maciej Główka said…
In my opinion. autonomous cars regulation is one of the biggest problem we have to face in near future. Autonomous cars will be among us quicker than many people think. In my opinion, car should protect driver and passengers. If something happened, who should be responsible for accident in such situation? If driver changed something in algorithm, that caused incident, then situation is clear. Otherwise, if it is clearly car fault, then I think company should be responsible for it.
All in all, I think that autonomous cars will decrease dramatically number of car accidents.
Vladlen Kyselov said…
I think that if we will have autonomous cars in our roads we will have to make separate roads for them or remove human driven cars at all to reduce all risks. Also it has to be well tested system, because it will be responsible for lots of humans lives. I guess it will definitely reduce car accidents, maybe there won`t be such accidents any more. I am absolutely confident in that it will solve the traffic jams problem as well. I really hope that this will happen in the nearest future. However, I guess it should have a function to make human take car under control but also with certain rules like being sober, adult and others.
Who would buy in his mind a car that can actually prefer life of other people more than yours? Who will be responsible for an accident? It may sound harsh and selfish, but i am not eager to sacrifice my life for random people that i don't know, because they are running through the road on the red light or car just malfunctioned.
Unknown said…
I can say that I am a more traditional driver. I prefer a car where there is no new technology, such as parking sensors or a parking program. Of course, there are systems that are important such as power steering or ABS. I like to feel a car, to decide for myself how and where.
I do not believe in autonomous cars. For me they have too many hidden bombs. A computer will never replace a human being. Who will be responsible for the accidents?
In my opinion, this is a mistake of today's science, which is heading in the wrong direction.
I have seen a lot of similar videos and there is still no answer to this question.

Even though I like driving on my own, I wouldn’t like to pay money for something that could potentially kill me.That makes no sense.
But on the other hand the car should still make a decision who should be dead or alive. This means that such cars should somehow compare human lives. But how? Should it take in consideration people’s age, material status, sex, weight, height, grades in the school? What should be the criteria for comparison?
Unknown said…
I guess a small percentage of people would like to know that there car could potentially kill them in favor of somebody else. I would chose to have some kind of survey on my personal preferences and ethical choices so car would do same as I would do in such a situation. And we for sure should integrate autonomous cars into our traffic due estimation if 10% of vehicles on the road would be autonomous car that will illuminate the most severe traffic jams due to irrational behavior of human drivers on the road.
Unknown said…
This is difficult problem because driver and pedestrians are important. But when we look at cars today and car companies, these companies mainly focus on driver, not on other people. We have few changes in cars for other pedestrians for example if you remember Mercedes sign on the front of car which was sticking out. Now cars can’t have this kind of signs because it was dangerous for people when accident happened.
Iman Masjedi said…
In my opinion, the use of technology, especially in the case of autonomous cars, reduces the number of human-caused deaths caused by accidents.
The cause of most accidents is human faults. Alcohol consumption or fatigue caused by sleepiness during driving, instantaneous thrills during driving or stuttering at speed.
The autonomous cars does not have the momentary human and emotional behaviors and can take the best logic of self-control At the time of the accident.
Unknown said…
I agree with the speaker that autonomous cars will significantly decrease number of car accident victims. However, regulations are of the great importance here because autonomous cars technology can lead to partial transfer of responsibility for accidents from drivers to car manufacturers.. I agree with the idea, mentioned few posts above, of designing some kind of common protocol which would be implemented by every autonomous vehicle and with principles which would be clear to everyone. This would allow for some kind of cooperation between vehicles and makes their behavior more predictable for humans.
When it comes to scenario which author has mentioned, I think that vehicle should then calculate the probability of death or serious injuries of passengers based on speed and other conditions and if it would be below some threshold it should decide to hit the car in front of it, otherwise it should of course swerve off.
I think that autonomous cars is our future and we need to go to it ASAP because this will solve huge amount of problems caused by "Human factor" and world will be more safer for us.
Illia Shynder said…
These questions are hard to answer, and I think until the moment humanity decides how autonomous cars should behave, until humanity solves this philosophical problem, autonomous cars shouldn't be implemented, because we can't tell robot what to do if we don't know what to do ourselves. I guess the good decision would be to try to make all-earth survey about how such cars should behave. Or to try really find a new way of solving this problem, like make some technology that will make humans inside the car absolutely protected from any damage(i don't know, maybe 100 airbags), so that car won't have to decide whether to kill owner and save more people or not. I'm not talking about only this particular scenario, but I think such technology would solve all or almost all such complications. It would be easier to make decisions if we can be at least 99% sure that human inside the car will be protected by some super advanced autonomous technology
I agree in 100%. Most accidents happens by human stupidity. If people were driving safely, not speaking on phone while driving, using turnsignals, watching pedestrants there would be like 90% less accidents. And if it comes to accident that car cannot avoid i think there will always be one option that "can work" i think car should choose this option if its also come with less casualities.
Unknown said…
I think that it raises an interesting topic of security. Unfortunately, I don't agree with what he says. Of course, you have to do a lot of testing before releasing it into the world. But we shouldn't be afraid that something can happen. It will always be like that, someone will always suffer but it should not be something that blocks development. Probably there will be accidents in these machines, but you have to learn from mistakes and correct them. Of course, not every accident is AI's fault. We need to develop this technology, this is my opinion.
Illia Lukisha said…
Maybe I'm a little bit cruel and pragmatic, but first of all, if death cannot be eliminated, car should decide who breaks the rules and then look for minimizing the risk, because if family of 4 people drive on opposite lane and bicyclist don't brake the rules than car cannot sacrify bicyclist even when the damage will be less.
That's my opinion.
Unknown said…
First of all the autonomous cars should do everything to prevent the accidents. Then it should do everything to minimize all costs from it. And if it's not possible, it should prevent the passengers, but may not do it without little harm for them. In most cases it would be probably the human mistake or some other kind of unpredictable error, that would lead to the crash. But it's the first generation of autonomous cars, so it won't be perfect. And it's hard to predict, how would it behave in other cases, not in that insulated scheme, because it may calculate really strange or amazing behavior, that will solve the problem.
Unknown said…
In my opinion in case when robot could not save everybody should not take action to choose and rank people lives. On the movie that was the opinion of the second quote philosopher which I agree. In that scenario will be less objection of people or families of the people that was chosen to dead. Computer should try to minimalize effects of the accident but should not change course to kill for instance smaller group of people. Everybody should check their course before action. And if that course is collision course then will be consequences.
Kristina Moroz said…
I always thought that the car should be safe both for people who in it and others. But when it comes to autonomous cars and people who use it, they must be aware of all risks that this technologies bear and be ready to be responsible for this.
Jakub Lisicki said…
Such a range of moral dilemmas is pretty tough to solve even for us - humans. We live in a society that is used to the fact that we can blame every fault on people responsible for them. Who should we blane then, when we take out the human factor out of the equation? Humans are not machines and follow their instincts during the most critical situations. Machines however, are able to analize everything and make a decision a lot faster than us. How should they react then? I don't think that anyone has an answer to this question, but I guess that no one would buy a car which wouldn't put safety of its' passengers ahead of everyone else's safety.
Lukasz Mroczek said…
Your right. People are not conscious about this fact but autonomous cars are more and more popular. That's why we have to take into consideration all facts and regulations and try to make it better
Lukasz Mroczek said…
I think that options you've mentioned are impossible and won't take place. But there comes a question how to test it properly and also what the outcome should be? What testers will be supposed to expect?
Lukasz Mroczek said…
I agree with you and I will never buy autonomus car because driving is my passion. I'm only a little bit scared about this cars because as you've mentioned computer will never replace human.
Lukasz Mroczek said…
That's the biggest problem that there seems to be no solution, no answer but we have to find it because this cars are becoming more casual and affordable for almost everyone now.
Lukasz Mroczek said…
To be honest i think that survey won't be good clue to determine how to write autonomous car alghoritm.
Lukasz Mroczek said…
There are some things that may somehow try to protect pedestrians but the problem seems to be a little bit different.
Unknown said…
As far as I'm concerned, this is very complex dilemma - but You can look at it in a simple way, maybe my thinking is wrong, but lets answer this question - "how much time, do You have during accident?" Miliseconds? Avarage driver don't even starts to think about the action and consequences of his choice - he just want to survive, hit the braks or dodge the car in front of him - Dodgin is also not a typical behaviour, I saw so many accidents, people just hit the breaks and counting their best... In my opinion this topic is a good base to think about this problem, there are so many variables and random events during accident so it's hard to invent an extrordinary algorithm - I will agree here, but..
All the regulations an street was invented so even the dumbest person, who passed the test and get their license can drive on the street. Speed limits was created so the driver can easly manage their car. 99% of the accident is because of people fault, people get distracted and lose control of the situation. But I'm not going to dive deeper into this topic, my point is - are we going to trust autonomous cars and eliminate human fault(like distraction) or stop this idea because we don't know how the car would react during accident? Like I said, regulations, laws and limits are made, so the idiot can stick to it and feel safe, I think, that if autonomous cars will stick to the limits, the probability of serious and dangerous accident should be low
Unknown said…
This comment has been removed by the author.
Unknown said…
For me the most interesting thing about autonomous cars is a way how they make decisions. I mean one simple situation when we are on the board of such car and there is a choice between save drivers/passenger life or life of a passers or other cars driver.
Few weeks ago there was a fatal accident and even with those all sensors and security systems car can't avoid tragedy.
"Imagine that you are driving and another car in front of you brakes hard. It’s obvious that you will try to avoid hitting him but on your left side you have a car with a passenger without seatbelts fastened and on your right side you see a motorcyclist without a helmet."

I think that you missed the part where human cant even make such decision of who to "hurt", i think that we come into autonomous driving topics with wrong attitude. Those cars this technology enables us to to even have such decision to make , Is it simple one ? NO! it never is but at least now we have a choice! And we hate that choice, oh we do we fear thet we wont be able to use our usual excuses like "I lost control over the car", "I did not react fast enough" etc. We want to make devil out of autonomous cars just because we fear being liable for what we are doing even more.

Lets look at this from the other side what if there are no non autonomous drived cars on the road ? Who is there to breake uncontrollably ? Who is there to cause an emergency on the road? People dont like this question because answer to that is "other people".
Unknown said…
I agree with the speaker that it’s more a moral choice for autonomous car AI. However, from my point of view the worst choice which we are doing now it’s that we are not allowing driverless cars for public use. Due to in any scenario inspired by Emanuel Bentham or Immanuel Kant when AI will need to make such moral choice the number of victims of driverless cars will be much less than by a car driven by a human. And the choice to not allow the driverless car on the road is resulting in that 1.2 million people die in car accidents each year.
Marcin Górski said…
It's very interesting topic.
I'm not sure that autonomous cars can't prevent all of that kind of accidents. As you said, they include a lot of algorithms which can have errors. In my opinion, computer can't replace human being.
On the other hand, there are a lot of too confident drivers on the street who are driving very dangerous. The same situation is with older people whose reflex is very slow. That's why I have thumps up for autonomous cars.
In my opinion, in the presented situation, an autonomous car should take care primarily of the passenger's good and safety. It is only then to decide on the safety of others. This topic is very ethically complex. Depending on where in this situation you are, the opinion about how the car should behave changes. It is natural because the individual will always choose his own good over the good of others. I believe that the use of such cars will be possible on a large scale, but more tests and tests should be carried out before this happens A few years ago I witnessed a situation when the driver turned sharply so as not to pass the dog and went to a woman on the sidewalk. He saved the dog, but it damaged the passer-by. I hope that such cars will minimize the damage in such situations.
Unknown said…
its an idea of perfect future to remove human factor from driving process, can't disagree with the speaker, but as for me these cars are not one hundred percent safety our-days, they have and the will be tested in much bigger amount of different cases and situations to be all over around us. I hope it will becomeMuch earlier than 2030 an ordinary thing much earlier than 2030. The future is now!
It is possible that money (which is not known for its moral focus) will decide, at least in the beginning. The most expensive cars, which already have the best crash and safety standards and options available, will be designed to protect their human passengers above all else. The least expensive cars and some pedestrians will be somewhat lower priority.
Peter Clemenza said…
Autonomous car should take care primarily of the passenger's good and safety - because no one ever would like to buy such car that during dangerous situations will choose to save life of people on the road.
Maciej Nowak said…
It is very complex dilemma and there won't be just one true answer to this problem.I don't like that everyone forgets that they can be both passenger of an autonomous car or a pedestrian depending on situation. If you think that your car should protect your life at all cost, remember that you can die because of some other person's car that will try to protect life of his owner. For me autonomous cars should minimise a loss in an accident. However, there is still question how to calculate which live is less or more important.
Foodocado said…
The topic was discussed many times. I believe the cars should consider all options and chose the best one. Even though the owner of the car could be sacrificed. We should remember that the autonomous cars are something new to us. Not everything is perfect at the beginning. For sure, these cars will have some supporters and haters, but in my opinion, we should give them some time.
Unknown said…
Speaker has mentioned Asimov's laws of robots. That's interesting because for me as long as there were only three laws everything was simple. Adding the fourth one brings the same social dilemma to the robots' world: what's more important: one life of life of sb else. I think that I'm not ready to answer for that question. It's connected with our believes and can't be simply solved.
Anyway: introducing driverless cars (even with principle to save the owner for all cost) will reduce accidents. So I support this idea.
Unknown said…
I think that even now, with nowadays AI level of autonomous cars, computer is better than human in making such decisions. In this kind of situations people behavior depends mostly on their emotions. Computer make decision on calculation and in all cases would choose the best and least harmful decision.
Unknown said…
When I think of any car accident I had, it comes to my mind that I was in shock, clearly wasn’t thinking what I’m supposed to do in such situation. It was all about reflex and driving skills, so when I think about that I could have a car with extra pair of eyes, always sharp mind, focused in every second on the road, never tired I say yes. I think in 99.9 % of cases the response is quicker and more precise than I could ever get, so I think the autonomous cars are the best option for people with low driving skills, disabilities and other things wich make them more dangerous than a computer controlled car.
Unknown said…
This video showed me the dilemma I wasn't aware. It's much more complicated than I thought. After seeing this and the results of the surveys they've prepared, I suppose it might take next tens of years to find an answer to these questions that won't cause any protests.
I hope we will be able to maximise the optimisations and do the best we can to prevent human society but as mentioned in the movie above we must agree on some tradeoffs to be able to cover all cases.
Lukasz Mroczek said…
Hmm. Maybe we shouldn't be afraid but we have to keep it in our minds and be conscious.
Lukasz Mroczek said…
I think that I agree with you ;)

Popular posts from this blog

Week 11 [03-09.06.2019] The problem with ecological cars emission in UK

The problem with ecological cars emission in UK Since the adoption of the European Emission Allowance Directive in the European Parliament, all car makers have tried to submit. Since 1992, the Euro I standard has been in force, which limited the emission of carbon monoxide to the atmosphere. The Euro VI standard currently applies, which limits the series of exhaust gases. These include: hydrocarbons, nitrogen and carbon oxides, and dust.   The most significant change was brought by the Euro IV standard. For the first time it introduced the limitation of nitrogen oxides, which are responsible for the harmful compounds of smog.   What is smog?   Smog consists of sulfur oxides, nitrogen and carbon. In addition, solid substances such as suspended dust (PM). Dust suspend in atmospheric aerosols may be in liquid and solid form. These can be particles of sea salt, clouds from the Sahara and artificial compounds made by people. These compounds often come fr

Week 12 (12.01-18.01.15) Are you an early bird or a night owl ?

Owls are nocturnal creatures. They’re wide awake at night and they sleep during the day. If this sounds like bliss to you, then, like about 20 percent of the population who find themselves most active at around 9 pm, you may fall into the same category as our feathered friend. Night owls often have difficulty waking up in the morning, and like to be up late at night.  Studies of animal behaviour indicate that being a night owl may actually be built into some people’s genes. This would explain why those late-to-bed, late-to-rise people find it so difficult to change their behaviour. The trouble for night owls is that they just have to be at places such as work and school far too early. This is when the alarm clock becomes the night owl’s most important survival tool. Experts say that one way for a night owl to beat their dependence on their alarm clocks is to sleep with the curtains open. The Theory is that if they do so, the morning sunlight will awaken them gently and naturally.

Week 4 [06-12.11.2017] This is what happens when you reply to spam email.

James Veitch is a British comedian. In today’s Ted Talk James with characteristic for himself a sense of humor shows how he deals with spam emails and why responding to junk messages may be sometimes dangerous. Questions: What do you think about James’s  way of dealing with spam? Why are junk messages legal, even though it sometimes may be a fraud? Dou you have a problem with spam? How do you deal with with it?