The harm principle is good in common cases, but I fear this may be an edge case, and the harm principle tends to break down when the meaning of “harm” or “hurt” is called into question. By the standards of Western Civilization, siphoning money from Joe’s bank account is harm to Joe, although any physical effect on Joe is very indirect; making out with someone of the same gender is not harm to Joe, even if the sight of it makes him violently ill. By the standards of Islam, drawing certain pictures can be harm to everyone of their faith.
It seems to me that there’s a narrow range of value congruence where the harm principle is applicable; go further and it is incoherent, closer and it is redundant.
I agree with this. “Harm” is too vague to make the harm principle a fully general argument for the Western liberal order- and it certainly wouldn’t do to try and program an AI with it. One thing a liberal society must wrestle with is what kinds of behavior are considered harmful. Usually, we define harm to include some behaviors beyond physical harm: like theft or slander. But watching computer generated images of any kind, in the privacy of your own home is pretty solidly in “doesn’t harm anyone” category, as defined by the liberal/libertarian tradition.
Part of my point is that there isn’t really much of an argument to be had. I suppose if someone demonstrated that the existence of computer generated snuff actually threatened our civilization or something, I could be swayed. But basically I think people should do things that make them happy so long as they avoid hurting others: if that isn’t a terminal value it is awfully close.
The harm principle is good in common cases, but I fear this may be an edge case, and the harm principle tends to break down when the meaning of “harm” or “hurt” is called into question. By the standards of Western Civilization, siphoning money from Joe’s bank account is harm to Joe, although any physical effect on Joe is very indirect; making out with someone of the same gender is not harm to Joe, even if the sight of it makes him violently ill. By the standards of Islam, drawing certain pictures can be harm to everyone of their faith.
It seems to me that there’s a narrow range of value congruence where the harm principle is applicable; go further and it is incoherent, closer and it is redundant.
I agree with this. “Harm” is too vague to make the harm principle a fully general argument for the Western liberal order- and it certainly wouldn’t do to try and program an AI with it. One thing a liberal society must wrestle with is what kinds of behavior are considered harmful. Usually, we define harm to include some behaviors beyond physical harm: like theft or slander. But watching computer generated images of any kind, in the privacy of your own home is pretty solidly in “doesn’t harm anyone” category, as defined by the liberal/libertarian tradition.
Part of my point is that there isn’t really much of an argument to be had. I suppose if someone demonstrated that the existence of computer generated snuff actually threatened our civilization or something, I could be swayed. But basically I think people should do things that make them happy so long as they avoid hurting others: if that isn’t a terminal value it is awfully close.