Hands-Free Driving

Release Hands, Are We Ready Yet?

Group 57

As the development of artificial intelligence and SLAM autonomous systems become extremely potential research projects for big companies. It seems that driverless vehicles will be the future of the vehicle industry if those companies enable to overcome technical challenges. However, it will conflict with the current law systems and ethical standard which needs a feasible solution.

Potential benefits of developing and popularising autonomous car

Currently, the features of each kind of existing cars are including cruise control, traction control and anti-lock braking (ABS), which represents the autonomy in cars having existed. These technologies have been very developed to intervene special cases when needed without driver control, aimed to make driving safer. In statistics, a total of 27130 people were killed or seriously injured (KSI casualties) in 2017.

Hence, why not improve it to the next stage by eliminating people driving mistakes and then usher in the brand-new period of fully autonomous cars to achieve fully safe driving?

The benefits of action of removing driver are not only reflected in the human’s safety but also embodied in environmental protection. The well-control of speeding up, brake and transmission give a higher fuel efficiency and a lower emission of greenhouse gases. In statistics, the fuel consumption rate and emission of relative gases would reduce 2~4% per years in next 10 years.

In addition, the time saved from driving can be transferred to work which extremely increases the productivity. With the combination of high productivity, reducing of a traffic accident and the growth of fuel utilization, it is able to bring values equivalent to 1300 billion dollars per year to US economy. The value benefits from this technology could be 5600 billion per year if it is promoted to global.

This technology will provide a chance for those alcoholic people as well, and it was reported that only 17% of adults didn’t consume alcohol or non-drink experience in 2015. With applying this technique to the future cars, it will greatly improve people’s life quality. They will be no worries about drunk driving. Driving for a long time is also another serious issue in the last few years. Thus, it is a fantastic idea to transform human beings from driver to passengers. As passengers, we can do anything we want during the driving process, such as phone calls, playing games, even working in the car. The most interesting thing is removing the requirement for driver controls, such as steering wheel, pedal plates, even accelerator, which results in more available space in the vehicles.

Autonomous ParkingWith the human population is dramatically increasing these days. In statistics, World population projected to reach 9.7 billion by 2050 which causes the high demand for living area. The promotion of autonomous vehicle can reduce the number of private cars so that the land built for the park can be saved to put into use of living. Even for a private autonomous vehicle, it would be led to the most proper parking area by GPS system. If it comes true, approximately 31 percent of land in a city can be transferred into a residential area.

Existing ethical problems developed inside the self-driving vehicle

There are no matching laws and regulations to identify who should be responsible for a traffic incident. If someone is killed due to the technical limitations of the driving system, who we should be blame, technical company or driver? Life is lost but no one will be sentenced just because of lack or even ignorance of the targeted law.

It was reported that a driver of a Tesla car had been killed in a road accident after its Autopilot mode failed to recognise an oncoming lorry. Recently, it is reported that an Uber autonomous car [9] hit and killed a woman in a new trial on the road. The self-system technology and corresponding law have not developed enough yet.

Let us dig deeper into potential accidents caused by limitations of this tech. If the autonomous car would decide to collide with trees or fence, it could cause hurt to the passenger in the car who has a clear mind just to avoid a crash with another car driven by a drunk driver.

For the most people, they agree that the man who has clear mind should not bear the physical pain and the damage to the private property. The drunk driver should pay the price both from economic and physical aspects. In common sense, people will protect themselves according to egoism, but programming logic of autonomous car could not handle the issue in this way. Although maybe the consequence of action results that they will all survive and protect the drunk driver who seems to be “weaker part” in this accident by sacrificing part of benefits of the others. It could raise the argument whether should save drunk driver’s life or not. This action violates the egoism and distinct but demonstrates virtue ethics, demonstrated in Kant’s Theory. Resulting consequence is central to the moral decision.

Technically, an autonomous system should not discriminate between individuals based on age, gender or any other factor. Autonomous systems do not have those human frailties and will choose actions via cold logic programming.

In most real situations, an autonomous car will check a value-of-life table and choose to hit the lowest value of life could be the best solution, but the ethical issues do exist in extreme cases. A model of human morality is very complexly, different cultures and regional groups may affect the results. It is disputable that which perspective of the population to use when creating the table. On the other hand, the sum of our ethical choices that transfer to vehicles could disclose humanity to be despicably racist and selfish. This is due to implicit bias, also known as implicit social cognition, which is pervasive and refers to the attitudes or stereotypes that affect our behaviour in an unconscious manner. This is a more harmful flaw that applying a controversial value-of-life system with our embarrassing ethical imperfection to autonomous vehicles.

54 thoughts on “Release Hands, Are We Ready Yet?

  1. “This action violates the egoism and distinct but demonstrates virtue ethics, demonstrated in Kant’s Theory.” Virtue Ethics and Kant’s Theory (which is also called Duty Ethics) are different. Alos can you say _how_ this action demonstrates Virtue ethics please?

    Overall, please expand on the ethical reasoning for and against.

  2. This was an intriguing read.
    I think that self driving cars provide many advantages and it would be a great technological development to have these vehicles commercialised. I like how the pro part was written and it made me think that it definitely would be better to have one of these.

    Personally, I think that whoever buys one of these cards, should be conscious about the problems that it could bring. There is the possibility of an accident happening and this person should be aware of it. If the driver is paying attention on the road, he/she should be able to take over the car’s control if he/she sees something that could cause an accident. I think that removing everything that doesn’t let the driver take over the car’s control isn’t a good idea and I don’t think this will happen in the future.

  3. In the future, the autonomous car must become one of the hottest topics in the car industry region. It also reflects some ethical problem for some special issues. For instance, as the article says, for drunk driving people, who needs to take responsibility for a car accident? The driver? or innocent passerby?
    This is really a thought-provoking question.

  4. From my perspective view, the autonomous car would be increasingly popular in future, but the safety must be the most controversial problem. Although Tesla car had killed a woman, it is safer compared with traditional Manual driving, because it has several different sensors and supervising all directions of surroundings. Meanwhile, there is only one accident during thousands of tests. Technically and statistically, the autonomous car is safer. If the autonomous car kills somebody, it always makes people feel like that technology kill somebody instead of an accident. It’s hard to make people change this idea.
    Elon Musk said, “If you’re sceptical of self-driving technology, you’re killing people”. Do you agree?

  5. I think the use of autonomous vehicles will grow rapidly in the future. However, the companies or the manufacturers still have to overcome the technological problems and ensure that the autonomous systems are reliable enough before entering the market. Also, the government should improve the current legal system as well, taking into account these autonomous cases.

  6. I am looking forward the life with self-driving vehicle. If the company could allow a self-driving car directly driving to my home everyday and pick up for going to work and school, we maybe have enough time to have a sleep and eat breakfast on the way to work and school. This tech could also reduce a traffic risk for some unpredictable issue, which could improve life safety and quality. Meanwhile, this tech also could reduce living cost for buying a car and maintaining it. I sincerely wish government could launch a law for protecting both owner and company’s right.

  7. A ethical problem indeed intrigue my thinking. Who to be blame for the properties even the human life if the traffic tragedy is cause by self-driving? I think the government and company should push the specific law launch together. If a customer buy a car from this company and also pass the driver examination, i think he or she should take this responsibility both from law and ethical aspect. He or she should be sentenced by the law if he or she ruin the other’s life and the compensation for money could cooperate with insurance company. These are my own points.

  8. The loss for self-driving technique is inevitable for the development. We just overstate this situation. Every year many lifes loss because the traffic issues and we don’t make any traffic to be a hot topic for us to discuss. Because this tech is totally new and novelty, people show their anxious and worried to it. That can be totally understand. This ethical problem should be empahsised but it is not nessary for us to stop this tech development. We can encourage the government to launch this law to protect both rights!

  9. A good paper. It has discussed deeply about the unpiloted technology. What’s more, it focuses on the ethical problems which is one of the main obstacles of this technology

  10. I am quite interested in self-driving vehicle. I dream for buying a this car like Tesla on the way. It is very cool and amazing tech. Imagine that if we go to self-driving tourism, we can use this car to the place we do not know and even we have not planned yet for. With this kind of car, it can bring us a lot fun and high standard protection. The ethical problem I am not worried about it because we even possessed car in the hundred year before, but we do have a complete law for traffic. I am confident for the development of self-driving car.

  11. I am really interested in the self-driving vehicle because I am the driver who not confident about my own driving skill. When I drive on the road I would feel nervious, which makes car accident happened more frequently. Withing the technological help, I believe that the rate of the car accident would be decresed, but we cannot trust the machine entirely. therefore, it is better to combine the human skill and machine together.

  12. For the foreseeable future, I believe that autopilot is almost certainly a substitute for human driving in most cases. Legal issues may be a potential obstacle for unmanned driving, especially with regard to the definition of liability for accidents. For example, how to define an unmanned vehicle accident is the responsibility of the owner or the issue of the producer. Driverlessness also faces ethical issues, such as the fact that in certain situations it is necessary to strike a person, and if you are confronted with A or B, you have to choose. Fortunately, at the legal level, it is the result of trade-off and compromises. This problem can always be agreed upon after the formulation/modification of rules, and it is not unresolved. However, the problem at the moral level is often not a problem, so the above problem is not a problem either.

  13. People’s concerns about the safety of unmanned vehicles are superfluous, and the data are always the most convincing. Even aircraft, an early transportation vehicle with high crash rates, has gained popularity in the modern era. With the advancement of technology, how can a more secure unmanned vehicle probably not be popularized? If Tesla did not lie, Tesla’s equipment also proved that the price will not be a hindrance factor. Tesla, Google, Uber, General Motors, Toyota, and Honda are all investing heavily in unmanned vehicles, and more car companies have changed the attitude of unmanned vehicles, from thinking that driverless technology is immature to promising with a great opportunity. This shows that the window has really been opened and a perfect opportunity has come.

  14. Interesting article. In my point of view, the self-driving car is still a valuable technology for the future. The aircraft already have autopilot fiction, why not the car? The autopilot are not perfectly developed, public need to give the technology more time to develop. I terms of responsibility, the autopilot car cannot drive without driver, and the driver still have the main responsibility for the accident as the driver agreed the terms and conditions when they using the autopilot. In conclusion, the autopilot still valuable and it is worth to develop.

  15. Interesting article about currently very imporatant issue.

    However, I believe that driverless cars is the future. The point about the deaths caused by autonomous vehicles is, in my opinion irrelevent, as media exposes these accidents and everyone thinks it is a big deal while thousands of auto accidents happen every day caused by careless drivers and it has become a natural thing.

    With an improvement in the technology, further generation autonomous vehicles are going to be even more safer, save fuel, time and will require no interventation. It is just a matter of time when goverments and law makers catch up and we will not be suprised by cars without a driver

  16. Compared to pros & cons, I am more intrigued by the ethical concerns. Technologies always take us to a new dilema while fitting the needs. The ‘value-of-life table’ is indeed a thought provoking point, I must say. Human morality is a complex of humanity and other factors like regions and genders, which can be controversial for even ourselves. Simply Transferring it to codes without thorough thinking will amplify the problems.

    For example, if the unmanned car killed a senior citizen to avoid hitting a kid because of its codes, wouldn’t the family of the victim reckons the company setting up the rules murderer? Finding a trade-off will never be easier than the technical difficulties.

  17. Exciting article. I don’t know how to drive so this is an excellent technique for me. It can also help the driver when the road situation is complicated. I think this system could be an assistant for the driver which means the car is semi auto, so the driver still has the main responsibility. Hope this technique will develop fast and will be available soon.

  18. As far as I’m concerned, self-driving cars will replace the drivers inevitably in the future. It is considered as the next generation of the driving technology. Although there is still a long way, the technology tendency is clear. I think people will make a good balance between different aspects. Especially with the expansion of AI, the security and stability of this technology will be improved gradually. Overall, I’m confident with self-driving cars.

  19. a good and interesting article. Personally I think autonomous vehicles are the future, not only the environmental impacts from GHG can be largely reduced, but will be quite efficient for the growing global population. Although currently no clear rules have been set for this area of technology, and many technical problems remain unsolved. It may not be realistic to put into large production right now.

  20. Personally, I believe that AV’s bring more ‘good’ than ‘bad’. There are really only benefits from implementing them, the statistics speak for themselves, however, there is no getting away from the fact that there are some serious ethical questions that need answering. One solution to the ‘who to kill’ dilemma is to not ‘choose’ to kill anyone but instead try to avoid all fatalities at whatever cost, the outcome would be left to fate, very similar to the reaction of most humans in the seat. The other difficulties are in the costly lawsuits that would inevitably ensue after an incident, I imagine that the blame would be on the car companies and not the customers as it was not in their hands. Overall, even though there are some difficult decisions to be made I believe that AV’s are a very positive step forward.

  21. Unmanned driving has a moral problem. If it is changed into a human driver, its own subconscious will make a choice, but it is a personal problem, especially in the face of law and morality, and the solution can not be solved by human beings. I believe that one-day unmanned driving technology from the laboratory to the mass market will bring new demands to the existing transportation system, the legal system, and the social security system, and launch new challenges. It will be a new topic for all sectors of society to make driverless technology faster, more stable and safer landing.

  22. I would like to see this car widely used. By taking this car, people can save time from that, especially those who work or live in big country. And decreasing of fuel comsompution saves money for car owners. However, I also worried accidents may happen, because the traffic can be really complex sometimes. The car may not avoid accidents like human. Anyway, hope to see more such self-driving cars.

  23. This is surely an intriguing article. The idea of letting robotic cars drive us humans around would surely be a step forward to a word more automatic and dependable on technology. Traditional people will surely have a problem maintaining the idea that “computers are taking over”. But the benefits stated in the article are surely something to take into consideration. The main purpose of the human from ancient times is to limit the pain and work that has to be done so he will have free time so this step serves that purpose.
    This article though states the question of how will the laws and the court respond to a machine making a mistake although it would be the most calculating move to limit mistakes. Robots make decisions after explicit judgment of the situation without emotions involved. So indeed how can humans under the guidance of the laws judge such a decision. Of course this question could be answered after decades or centuries when autonomous cars will already be a normal thing occurring in the society. Maybe in the future laws will shape themselves to also serve situations where autonomous cars would be in the court.

  24. Despite facts and figures suggesting that the introduction of autonomous vehicles would on a macro level benefit society, this is, in my opinion, challenged by the fact that as a population we have come to terms with the concept of ‘human error’ and the fact that humans make mistakes, yet struggle to be at peace with computer error as there ‘must’ be someone to blame that could have done more.
    In addition, the true benefits of autonomous cars will only be seen when ALL of the vehicles on the road are autonomous. Until then, autonomous cars will continuously be facing human drivers and using code to challenge human judgement.

  25. Is good to have the autonomous car to reduce accident due to human mistake. But as a driver we should full aware and focus to the road as the sensor might be malfunction or cannot detect the obstacle. for me, I still do not trust the capability of the autonomous car as the car still drive with other non-autonomous car. besides that, the autonomous car is very expensive such as tesla. I think full autonomous car is a cool invention because now everyone can drive. My concern is what will effect toward driver supporting sector such as taxi?

  26. This article is quite interesting but it has not developed deeply with four theories. But i think the future trend will concentrate on population of autonomous vehicle. The ethical debate is always involved with new technology which can truly change the life style of human beings. The action to moderate the wrong aspect in ethics should be decided by the public , government and the producers. The common duty could push the market the life to a better balanced point.

  27. Autonomous driving is the trend of the future and the product of continuous technological progress. But autonomous driving techniques can only be applied to people’s lives when they are mature. I think technology companies need to take responsibility for accidents to show their confidence in their technology.

  28. A good article. Yehn, autonomous vehicles quite have numerous benefits for people. But there exist plenty of moral problems as well if meeting traffic accidents. As the article says, who needs to take responsibility for those accidents? Driver? or company? And i think everyone saw the movie Fast & Furious 7. There also has some ethic problem for that, if using fully autonomous vehicles in the future. Hacker will easily control the traffic system, causing the traffic paralysis, even doing more harmful things. Have you thought about this issues if implement this technique into vehicles in the future?

  29. it is a big problem. It will take a extremely long time for self-driving car to replace private car totally because relevant policy should be carried out before self-driving is popularized and it is difficult to deal with the abandoned private car.

  30. The article presents in-depth information regarding one of the most popular aspect of automotive engineering these days, which is driveless vehicle technology. It covers various potential advantages that this technology is likely to achieve in the near future, while presenting several serious ethical issues emerging these days. Chief among which is whether a driveless car should decide what to crash into. If it decided to collide with an innocent person just to avoid some one more “important”, does that mean we are using technology to murder that person? Those issues are definitly reflecting the missing piece of our present mechinary of the law, and are something a responsible engineer should think about.

  31. I don’t think the development of automatic driving is a urge thing. Short-distance driving is not a tiring thing as far as i am concerned. I think nowadays’ automation should be used in other ways first and then apply it to vehicle when it is mature enough.

  32. I would say, auto-matic self-driving cars can gather “prior information” based on their “training rules”. These rules may refer to deep learning algorithms. The rules can get more and more sophisticated and basically capture the situation that happens most often. However, NOT ALL traffic accidents are due to systematic errors. Instead, many of them are due to randomness. Though randomness can be reflect in some way in the model, but some randomness are not predictable. Although there is good prospect for auto-matic self-driving cars in the future, they are intrinsically as dangerous as human-drivers. But they probably are sophisticated drivers.

  33. Users of self-driving cars are likely to worry that their personal information will be collected and used without their knowledge, purpose, or consequences for the users themselves. Self-driving cars essentially automatically collect and display data on when, where, and how a user moves from one place to another. Will users worry about the use of such personal data? Why collect these data? How will the data be used? How long will the data be saved? Who can access this data? Automobile manufacturers, fleet operators, and other automotive service providers (such as providers of car networking, map services, etc.) face the practical difficulty of obtaining user consent when vehicles are used not only by car owners but also by third parties. If most of the self-driving cars are operated by the fleet rather than the consumer alone, the problem will be complicated. Automakers and operators should also be vigilant about the risks posed by third-party providers that process data on their behalf. If the car manufacturer or operator cooperates with the technology company for the interconnection service, and the partner violates the data protection regulations, resulting in data loss or misuse, this may lead to responsibility or damage to reputation.

  34. Yes, considering privacy issues, legislative and regulatory guidelines will help solve problems brought about by new technologies. In many cases, legislation is the key reason to promote or hinder new technologies. For privacy protection, although it is unlikely that there will be a law that specifically regulates the privacy of self-driving cars in the near future, the UK government may issue regulatory guidelines to provide maximum compliance with privacy regulations for automobile manufacturers and other stakeholders within the current legal framework. guide. Respecting consumer privacy is an important concern for consumers. Establishing privacy protection mechanisms for self-driving cars from the outset is the most effective strategy. In other words, automakers and other parties need to consider privacy issues at design time to minimize the amount of personal information generated, collected, or retained by self-driving cars. Further technical safeguards are the encryption and anonymization of information and the prevention of abuse or unauthorized third-party access. If automobile manufacturers and other parties involved in the development of autonomous vehicles can follow the privacy protection guidelines issued by the Auto Manufacturers Alliance and the Global Automobile Manufacturers Association, this can further reduce consumer concerns.

  35. In the automatic driving mode, since the vehicle driving activity is controlled by the vehicle’s autonomous driving system rather than by humans, this makes the fault liability lose the applicable premise, and involves applying the product liability problem without fault liability. At this point, the following issues must be considered critically in the revision of laws. 1. Insurance system. The existing mandatory accident insurance liability and rate are the results of comprehensive consideration of the basic protection of the victim, the level of economic development and the ability of consumers to pay, the incidence of traffic accidents, and the corresponding cost-benefit of insurance companies. After the application of automatic driving technology, traffic Changes in accident rates, technological maturity and social affordability will inevitably affect the above-mentioned insurance coverage and rate factors, and the amount and rate of compulsory liability insurance should change. 2. The application of no-fault liability. The traffic accidents resulting from the automatic driving mode may be subject to the “Product Quality Law.” The manufacturer or seller of the vehicle or autopilot system shall bear no-fault liability. In addition, the “Product Quality Law” has “disclosure provisions for the existence of defects in the science and technology level when products are put into circulation.” However, this is the logic of non-artificial intelligence era, and the “black box” algorithm in artificial intelligence—that is, the process of deep learning of artificial intelligence is similar to the complex thinking process of the human brain, it is difficult to accurately determine its final result, whether the above “product quality can be applied”. The exemption clause in the Law that has not yet discovered a defect is an important legislative and judicial issue.

  36. A sufficiently good autopilot system may be safer than people driving. People won’t worry about accidents caused by drunk driving, tired driving, etc. And with the development of technology, the safety of automatic driving will also be greatly improved.

  37. A sufficiently good autopilot system may be safer than people driving. People won’t worry about accidents caused by drunk driving, tired driving, etc. And with the development of technology, the safety of automatic driving will also be greatly improved.

  38. Automatic driving has been controversial for years. While there definitely will be benefits and advantages brought by this kind of technology, problems still exist as barriers to develop and expand it. For example, who’s gonna be responsible for the accidents caused by automatic driving vehicles? How to measure the specific responsibility? Does those technology companies meet triple bottom lines? Is it really ethical? Even though having been invented for a few years, automatic driving still can’t be mainstream not only because the technology is still immature, but also the possible economic, and political consequences brought by this. A tremendous number of jobs would possibly lost, and political issues would occur because of ethical dilemmas. All of these questions remain to be solved.

  39. Artificial intelligence could bring a new industry to the world, however ,, in a different forms. Autonomous cars driven by AI sounds like an ideal solution to traffic problem in far future but is far from feasible in real society due to legislation and ethics. AI is not clever enough to judge issues involving human emotions, even it can, it will become the engineer who decide the temper of a car which is impossible since no one is allowed to use his own emotions to solve problems for others without considering their ideas when driving driverless cars.The most feasible approach to this issue is to keep those tons of steels riding inside laboratories, while implementing other vehicles with driving auxiliary system with very basic controls over the car such as stopping a car when sth is too close in front of it. That means authority of driving is still given back to humans since we could not let an engineer who is not at the scene to decide the survival of two drivers.

  40. It really raised my awareness of automated driving and its related ethical issued. Before I’ve long been taking it as a moral dilemma, as is mentioned here, for which we’d never get to know the best solution. But this provided me with another perspective on driver fatigue and drunk driving. And one could definitely argue how a large number of accidents get avoided with this. The environmental impact was also thought-provoking.
    It’s just that the essay could be better if more information on laws on automated driving is mentioned. Is there a gap between current practice and laws at present? Or have any countries got their own laws & regulations?

  41. There’s just one point I don’t really get, ‘the time saved from driving can be transferred to work which extremely increases the productivity’. If I’m right time won’t be saved because the driver would still be travelling by a vehicle and it actually takes the same time? The only time saved could be that you spend on learning to drive but I don’t think this would ‘extremely’ enhance production. Please tell me if I’m wrong.
    But generally it’s very good! That’s a really thoughtful piece of work! Quite enjoyed reading through and getting to understand the ethical issues.

  42. I would like to say it is impossible to allow vehicle to take over the control. People should be responsible for their behaviors and lives. I think engineers should invent a core logarithm with the combination of artificial intelligence and driver’s decisions. If the drivers don’t give a correct response to the accident, they should be blame for that and get legal judgment. If the engineers could invent this tech, the ethical dilemmas will not exist any more. But this tech would be difficulty to create and the extent for human decision should take account for is also difficulty for determining. I totally support the development of self-driving car and want to see its future.

  43. From the perspective of the entire industry, the better direction is to start from its own perspective and look at what the upstream and downstream services of the industrial chain can do under the background of new technology and new business, including insurance companies and the legislature. I believes that once the ethical dilemmas are resolved in these areas, the commercialization of driverless cars will be smoother.
    We must also note that at the technical level, when the driverless vehicle and the Internet of Vehicles are sufficiently developed, the probability of car accidents is lower than that of manned vehicles. This will not only make the insurance industry willing to make a difference in the field of driverless vehicles. The development of the entire industry mechanism towards a win-win model is also a key point for the popularity of driverless cars.

  44. The decision-making of a car can be based on three principles: One is the principle of “protectionism,” and the primary task of a car is to protect the safety of passengers. This is similar to the current practice of auto-driving car manufacturers; the second is the “humanism” principle, where cars try to save the most people; the third is the “priority of interest” principle, and cars will protect the most valuable property. In the simulated situation, human life is always threatened. Self-driving cars need to decide who survives and who will die. He wants to remind people that ethical decision making is an extremely complex issue that can hardly be reduced to an algorithm. When studying the moral issues of auto cars, machine ethics cannot exceed human morality. However, if you are thinking about machine ethics, you must first think about your own morality. Am I really better than a machine?

  45. The “Trolley problem” is a classic question of moral philosophy: If you say that you drive a train, there are two roads in front of you, and you have a switchboard at hand. If you do not pull, the train will go the same path and kill five people; if When you pull it, the train will take you to the road and kill you. Will you pull or not pull it?

    This is the core of the controversy over traditional moral philosophy: the distinction between utilitarianism and absolutism. Utilitarianism says: Of course, it is better to die a person than to die five people; absolutism says that the act of stumbling over you is equal to killing. A person who will never die, so no one is better than that! These controversies have continued for generations, and there have been countless different versions of trams. Until today’s era of autopilot, there are still people who are asking similar questions. But the question here is naturally: Is an absolute, abstract ethical issue applicable to specific situational analysis? Does the traditional moral philosophy apply to our judgment of specific issues? Perhaps we want to introduce modern cognitive science research conclusions.

    For a long time, moral philosophers in the classical sense have derived moral reality based on rationality. They believe that human beings judge moral issues through rationality, and utilitarianism or absolutism are arguments based on rationality. However, modern cognitive science tells us that human judgment of morality is not made through rationality at all.

  46. I am an image recognition engineer and I am very interested in this topic. People may often think that strong artificial intelligence as an artificial intelligence used in autopilot. It means that the choice of a self-driving car will take into account legal, moral, sympathetic, loyal, and other factors. But in the eyes of engineers, things are not like this. Artificial intelligence in unmanned driving is only weak artificial intelligence. It does not have such awareness of legal morals at all. In fact, the decision problem here can be regarded as a path planning problem. If it does not turn, it represents what kind of walking path a driverless car chooses. As for the traditional path planning problem, the process of decision-making that no one drives in considering all aspects of the current surrounding environment, laws and regulations, and even ethical and moral aspects are actually the expression of these various factors in mathematics. Under the framework of these factors, solve an optimization problem. That is to say, the decision made by the self-driving vehicle under weak artificial intelligence depends on how the model is built when the engineer implements the decision algorithm. If the driver’s safety is more important in the model, the car will crash into it; if the model optimizes the total number of casualties, and the car will burst into the river.

  47. First of all, the highest guideline for self-driving cars is to protect manufacturers.

    Specifically, it is such a few points:
    1. Maximize passenger safety. No one is willing to buy an unsafe car.
    2. Minimize the manufacturer’s legal liability in accidents.

    These two points are the same in most cases. However, there are inconsistencies. Take the example of “suddenly jump in a guy on the highway.” The decision logic of an autonomous driving system is not difficult to understand. For the following options:

    A. Choose to do its best to brake, do not change the path, and smash pedestrians – pedestrians should be responsible for violating traffic regulations. The manufacturer has no responsibility and the passengers are also protected.

    B. Choosing Sacrifice Owners and Passengers – “Choose Sacrifice” is legally equivalent to “Choose to Kill.” The manufacturer will definitely die in criminal and civil lawsuits. Who will buy a car that chooses to kill you?

    C. Choose to hit a third person – equal to the second one.

    From the market, no one will buy a car “Church of Our Lady” that may voluntarily sacrifice his own life. Manufacturers can never be so designed. However, the manufacturer’s argument is vague, and I remember that Mercedes-Benz’s answer to the auto-driving car ethics question is similar to “Our car is very responsible, and it will protect everyone’s safety” (without responsibility for vague references). Every accident in a self-driving car is bound to be thoroughly investigated, analyzed and announced. If it is revealed that a car will make decisions for “sacrificing passengers” under any circumstances, the company will close down the next day. As far as consumers are concerned, many people may be able to say it nicely. Statistics show that pedestrian safety and safety equipment for cars is the most neglected option for consumers. Obviously, nobody is interested in spending money on the safety of passers-by. This is only money, not to mention when it comes to your own life.

  48. In many virtual scenes that question the autopilot AI decision, there is no solution to a person as well. Once you move from an extreme case scenario to a widely applicable decision rule, the core is two perspectives:

    The first perspective is the rules, that is, traffic regulations, and the derivative rules based on traffic regulations. Modern and developed traffic rules have clearly defined the division of responsibilities and behaviours in various situations. Both humans and AI must strictly observe these rules. However, the current traffic rules are not designed for AI, and there are also deficiencies in it, so it is feasible to expand and refine it on the basis of the current situation as an underlying strategy for autonomous driving.

    The second perspective is efficiency judgment, which is based on not violating laws and regulations and corresponding rules, assessing the probabilities and losses that various behaviours may produce results, and choosing the behaviour with the least loss. This requires a quick analysis of the surrounding environment and the calculation of the possible consequences. In the final decision, the time may only be a few seconds, which very dependent on the ability of autopilot AI.

    In fact, the above two perspectives are also human’s own choice process under rationality: obey the rules, avoid the harm to the greatest extent, and when the injury is unavoidable, give priority to protect the person who obeys the rules and reduce the overall harm. This is also the optimal solution under the rules of human society. However, humans may make more emotional moral judgments. At the same time, the behaviour is also limited by the level of the driver. The situations are even more complicated, while AI can maintain absolute rationality.

  49. Pingback: porn movie

Leave a Reply