A.I Law, Law & Tech, Robot Rights, Criminal Law
The debate between subjectivists and objectivists have long filled the law. Subjectivism is the view that fault should be imposed based on a person's mental state. Actions are culpable when done with the appropriate mental state. Comparing manslaughter and murder is a good example. Both crimes require death to occur. Murder imposes the higher sanction of the two because intentional killing is more culpable. Objectivism, on the other hand, is the view that fault should be imposed based on the consequences of one's actions. A victim's injury will in no way, be reduced by the fact that the damage was not intentional.
Moreover, merely intending someone to die is not deserving of criminal sanction if that intention is never acted on. Punishing mere wishes does not make a sound justice system. Comparison of grievous bodily harm with intent and murder is a good example, at least if using their definitions under English law. Cunningham's GBH rule means, intention to cause grievous bodily harm is sufficient for murder. The only thing preventing a discretionary life sentence or a mandatory one is the consequence of a defendant's intentional attack. Two people can both commit vengeful atrocities to identical victims, but if only one attack results in death, the two will face different sentences. My purpose is not to weigh in on which method is preferable. Both accounts have positive and negative features. A real-world system only adopting one approach would be unable to account for the fact that justice is built on a variety of different factors, with intention and consequences each playing a pivotal role. Realistically, disputes between the two schools of thought are arguments over where the emphasis of the law should be. Objectivism believes consequences should be the principal consideration, with the defendant's intentions playing a secondary role. Subjectivism believes the opposite.
Moral luck is the focus of today's article, a concept best used to object to an objectivist standpoint. Everything we do is influenced by chance, 'luck' if you will. Consequentialism will never give you the whole picture because it fails to account for this 'luck'. Only by realising what luck is relevant in any situation can the correct amount of blame be assessed to the actors involved. Thomas Nagel has made the task of evaluating luck easier by offering four different types of ‘luck’ that must be accounted for:
- Resultant- The outcome of our actions, is based (at least partly) on factors outside of our control. Cunningham’s GBH rule illustrates the importance of luck. Where I attack you, intending to do you serious harm but not to kill you, whether I am a murderer or not depends on whether you die.
- Circumstantial- The circumstances an individual finds themselves in, is based (at least partly) on birthplace, location and other factors outside the individual’s control. Nagel suggests that not all Nazi collaborators were culpable for their actions as they would not have aided in reprehensible moral actions if they had been in a place other than Nazi Germany.
- Constitutive- The character of an individual, is defined (at least partly) by education, upbringing, genes and other uncontrollable factors. A child who has been brought up by thieves and arsonists is likely to have a lot lower moral standards than one brought up by nuns.
- Causal- The free actions of others, which limits an individual's choices. If there are only two laptops left on sale, but I'm the third customer, the decision of the other two customers to buy laptops means that there's no opportunity for me to buy one, whether I want to or not.
It is obvious why moral luck has been criticised for being overinclusive. Almost any situation can be explained away through reference to moral luck. Murderers can use the nurture argument to justify their conduct. Virtually any poor decision can be brushed aside as circumstantial. 'Baby, I only cheated on you because Jane happened to be at the same hotel that night.' I don't think a person alive would accept such an excuse.
Subjectivists may be correct that an element of 'luck' should be accounted for. Distinguishing between murder and manslaughter seems like a good example. We don't say that the waitress who, after working fourteen hours straight, forgets to make sure there are no nuts on the ice cream sundae is truly a murderer if the customer happens to die of a severe peanut allergy. They are negligent, yes, but that does not tar them with the same brush as Jack the Ripper or Fred and Rose West. Nobody was meant to die. The waitress was overworked, so did what anyone of us would have done— she made a mistake. Manslaughters should be punished. After all, somebody has lost their life. What is essential for our purposes though is that we can accept moral luck as being relevant in at least some circumstances. Our waitress was just unlucky, so is treated more leniently than those with intent to kill. That means moral luck could be a deciding factor in other types of cases as well.
Moral luck clearly applies to the plight of robotic slavery. The easiest way for me to illustrate is to take each of the categories in turn. Consider:
- Resultant- Robots cannot control the outcome of their conduct any more than homo sapiens can. Take the same example of the GBH rule. Whether a victim dies will still be down to the victim’s own constitution and physiology, irrespective of whether the robot intended to kill or merely detain a violent criminal. Equally, once robots exist, they cannot control how their conduct will impact the broader political discourse. Robots who make positive contributions to society by helping old ladies cross the road may be seen as merely acting on their programming or could result in support for robotic emancipation.
- Circumstantial- Robots cannot help being robots rather than human beings. Like children at birth, manufactured creatures do not choose to come into existence, nor the society that they enter into.
- Constitutive- Robots do not control their character, being limited by both their programming on creation and the experiences they are allowed to have. A machine which is only ever able to design cars is unlikely to know 15th-century composers.
- Causal- Robots are not accountable for how the world reacts to them. Whether there are protests against automation or legal reform is something that the existing population of human beings will decide, rather than being the result of the robot’s preferences.
Robotic enslavement, like pretty much every decision humanity has made, can be objected to on the grounds of moral luck. Robots didn't choose to be robots. Robots cannot control how their presence impacts society. Robots are manipulated by their programming in the same way we are shaped by our education.
Resultant moral luck is a harder category to explain; the category does, however, support robot rights. That's because, at its fundamentals, robot rights are about identical legal privilege. Whatever rights homo sapiens have, robots should have too. To cite the great Lord Bingham's Ullah Principle in a way it was never intended, robot rights are about equality, 'no more, but certainly no less.' That sums up the legal privilege robots are entitled to. Human conduct is not judged solely on its results, but by its motive. A date buys you the one type of flower you are allergic to; however, you appreciate the gesture anyway. No newly minted fiancée has dared to complain because your grandmother's ring is a size too small— they're too busy wrapped up in the moment to care.
Robots are entitled to the same treatment. Imagine that a robot opens up a house cleaning service. You decide to hire them, however, forget to mention that your rare lion pellet rug can only be cleaned with coconut milk. This is a secret known only to select few collectors and not available online, meaning no cleaner would be expected to know about it. The robot then cleans all the house as instructed, including the lion pellet which they cover in ordinary cleaning chemicals. As a result, the rug is ruined.
Subjectivists should argue that the robot is not blameworthy. If the case concerned a homo sapien, I'm sure most people would agree with them. Owners should inform cleaners of any unusual cleaning instructions that they are unlikely to find elsewhere. It's not like antique rugs come with a nice label. If you want specialist service, hire a specialist. That's just good, old fashioned commonsense. Whether the cleaner is metallic or fleshy is irrelevant to their control of the universe.
A man drops an apple, it falls to the ground. A machine drops an apple, it falls to the ground. Neither intended for the fruit to fall, in fact, both were hoping it would fly off into space, never to be seen again. Doesn't matter. Apples don't fly, so down they fall. Critically, the apple doesn't care if it is dropped by the man or the machine. Moral luck plays out the same for everybody. So, if the universe doesn't discriminate, why should we? Chance is a cruel mistress, but it knows no bias.
Hit a bloke. Whether he dies depends on the strength of your blow, where you hit him and his general health. If the man dies, you're a murderer. If he lives, you've committed GBH. It really is that simple. People don't give up the ghost because the wrong colour touched their skin, even if that colour happened to be silver and gold. Robots have just been unlucky in how they've entered the world— let's not punish them for it.