Westworld

I don't think I'm giving too much away if I say that the HBO series Westworld raises ethical questions about artificial intelligence. The series is set in an amusement park where people can live out fantasies of the Wild West, fantasies that to a striking extent involve murder and rape. But no one actually gets murdered, because the characters that populate the world are (very lifelike) robots. They feel no real pain, despite appearances to the contrary, and can be repaired relatively easily. And the customers are paying high prices for the privilege, so what could they have to feel guilty about?

I assume no one believes Jesus' idea that committing a sin in one's heart is just as bad as committing it in the flesh, but is also seems about as clear as can be that much of what Westworld's clients are paying to do is very bad (even though it is, in a sense, not really done in the flesh). Standard ethical theories seem incapable of handling this fact. Or, if not incapable, not at all well positioned to do so plausibly or simply. The problem is similar to the well known one about Kantian ethics and mistreatment of animals: if ethics is all about respect for reason and the creatures that embody it, then (why) is tormenting animals wrong? The Kantian answer is that it is bad because it makes tormenting people more likely, but this is fairly plainly inadequate. A dying man could spend his last moments torturing bunnies and do nothing unethical at all on this view. 

Shooting at (robots that look and behave just like) people for fun is cruel. Perhaps part of why cruelty is bad has to do with its effects in the world, but, as Kant saw, bad will is bad on its own, regardless of whether it turns out to have bad consequences or not in any particular instance. It seems to me that Westworld therefore shows that consequentialism and textbook Kantianism are wrong, or incomplete, as moral theories. I'm sure others see it differently though.

On a related note, in Avengers: Infinity War the baddie is a consequentialist and the goodies "don't trade lives" (cf. Kant and Romans 3:8). I usually don't like science fiction-inspired philosophy, partly because it's usually metaphysics or epistemology (which are not my thing) and partly because it so often seems to be wrapped up in concerns about what is cool (also not my thing). But Westworld, which also raises metaphysical questions, seems to me to demonstrate something important about ethics that is rarely shown.

Attempts to anticipate dissent:

1. The clients of Westworld don't do anything wrong--you can do what you like to robots, as long as no actual people's rights are violated.  In a way I agree with this. But perhaps that just shows that there is more to ethics than questions about actions and their rightness or wrongness. There is, it seems to me, just something obviously very bad about choosing to have the experience as of shooting a man and seeing him bleed to death, screaming in pain, etc. Perhaps the badness is located more in the heart-mind of the person choosing to behave this way, or in the choice to act this way, than in the act itself, but that there is badness there seems about as plain as it could be.

2. So you're saying it's wrong or bad to play video games that involve shooting, etc.? Not necessarily. But there is something bad about playing a game that focuses on violence and in which the players want the violence to be as realistic as possible. Space Invaders is not like this. Nor is Angry Birds. (Although would the Buddha play either of those games?) No doubt there is a gray area somewhere. Such is life.

Forsberg on Agam-Segal and Dain

Niklas Forsberg has helpfully posted a non-final draft of his review of Wittgenstein's Moral Thought (eds. Reshef Agam-Segal and Edmund Dain) here. Naturally you want to know mostly what he says about my paper, so here you are:

[I]t is a relief of sorts that only two of the papers of this volume start off by noticing that Wittgenstein wrote very little about ethics. [I'm pretty sure mine is one of the two.] One may even say that one of the most central lessons one should learn from this book is that “what we had thought of as the field of ethics is so vast and unbounded that we no longer recognize it as a field at all” as Duncan Richter formulates it.

[...]

Richter’s text is something like a turning point of the book. He focuses on later Wittgenstein’s more fluid conception of language, which enables him to bring into view the multifarious ways in which evaluative words – like good and beautiful – are meaningfully used. One may see this as on a par Wittgenstein’s earlier thought, where ethics is not about a specific something, and ethical difficulties may surface anywhere and are interwoven with our lives in language. If we assume that “good” means one thing, and one thing only, we will have nothing to say about it. As a consequence, a notion like “good” must remain indefinable, but not because it is ineffable or somehow out of reach. As Richter says: “An accurate  picture will not be neat,  because the use of the word ‘good’ is not neat” (154), falling back upon Wittgenstein’s idea that one cannot really “sketch a sharply defined picture ‘corresponding’ to a blurred one” (PI,77). We need to think about more things than moral code words such as good, evil, right, wrong, etc. One must also think about “things such as people, their property, our relationships with them, and so on. (…) There is far more to ethics than questions about what is right and what is good” (168).

[...]

[T]his is a very helpful book. It should be, and probably will be, one of those books that most philosophers who think about ethics after Wittgenstein will have to read.

Fixing Frege

Frege ("Logic in Mathematics"):

Definitions proper must be distinguished from elucidations [Erläuterungen]. In the first stages of any discipline we cannot avoid the use of ordinary words. But these words are, for the most part, not really appropriate for scientific purposes, because they are not precise enough and fluctuate in their use. Science needs technical terms that have precise and fixed Bedeutungen, and in order to come to an understanding about these Bedeutungen and exclude possible misunderstandings, we provide elucidations. Of course in so doing we have again to use ordinary words, and these may display defects similar to those which the elucidations are intended to remove. So it seems that we shall then have to provide further elucidations. Theoretically one will never really achieve one’s goal in this way. In practice, however, we do manage to come to an understanding about the Bedeutungen of words. Of course we have to be able to count on a meeting of minds, on others’ guessing what we have in mind. 

Also Frege (The Foundations of Arithmetic):

At first, indeed, [Mill] seems to mean to base the science, like Leibniz, on definitions, since he defines the individual numbers in the same way as Leibniz; but this spark of sound sense is no sooner lit than it is extinguished, thanks to his preconception that all knowledge is empirical. He informs us, in fact, that these definitions are not definitions in the logical sense; not only do they fix the meaning of a term, but they also assert along with it an observed matter of fact. But what in the world can be the observed fact, or the physical fact (to use another of Mill’s expressions), which is asserted in the definition of the number 777,864? Of all the whole wealth of physical facts in his apocalypse, Mill names for us only a solitary one, the one which he holds is asserted in the definition of the number 3. It consists, according to him, in this, that collections of objects exist, which while they impress upon the senses thus, \, may be separated into two parts, thus, (.. .). What a mercy, then, that not everything in the world is nailed down; for if it were, we should not be able to bring off this separation, and 2+1 would not be 3! What a pity that Mill did not also illustrate the physical facts underlying the numbers 0 and 1!

I'm no philosopher of mathematics, but Frege's criticism of Mill does seem on target. Mill goes wrong by imagining too limited a set of examples and by thinking of the one example he does consider as if it were fixed. But that is not the real world, and somehow we do manage to get by even though not everything is nailed down. But then Frege seems to want to pin down the meanings of words, despite conceding that in practice this is not really necessary and that in theory it is impossible.

Mill makes what, after reading Frege's criticism of it, seems like a stupid mistake. Frege's problem is of a different kind. There is something wrong with what he wants. He sees the problems himself, but still, apparently, goes on wanting the same thing. So pointing out the problems won't help at all. We might say he needs a kind of therapy, although this won't be regular psycho-therapy. Nor does it at all follow that therapy is what Mill needs.