Do you have a short summary of why he thinks that someone answering the question of “would you have preferred to die right after child birth?” with “No?” is not strong evidence that they should have been born?
I don’t know what Benatar’s response to this is, but—consider this comment by Eliezer in a discussion of the Repugnant Conclusion:
“Barely worth living” can mean that, if you’re already alive and don’t want to die, your life is almost but not quite horrible enough that you would rather commit suicide than endure. But if you’re told that somebody like this exists, it is sad news that you want to hear as little as possible. You may not want to kill them, but you also wouldn’t have that child if you were told that was what your child’s life would be like.
As a more extreme version, suppose that we could create arbitrary minds, and chose to create one which, for its entire existence, experienced immense suffering which it wanted to stop. Say that it experienced the equivalent of being burned with a hot iron, for every second of its existence, and never got used to it. Yet, when asked whether it wanted to die, or would have preferred to die right after it was born, we’d design it in such a way that it would consider death even worse and respond “no”. Yet it seems obvious to me that it outputting this response is not a compelling reason to create such a mind.
If people already exist, then there are lots of strong reasons about respecting people’s autonomy etc. for why we should respect their desire to continue existing. But if we’re making the decision about what kinds of minds should come to existence, those reasons don’t seem to be particularly compelling. Especially not since we can construct situations in which we could create a mind that preferred to exist, but where it nonetheless seems immoral to create it.
You can of course reasonably argue that whether a mind should exist, depends on whether they would want to exist and some additional criteria about e.g. how happy they would be. Then if we really could create arbitrary minds, then we might as well (and should) create ones that were happy and preferred to exist, as opposed to ones which were unhappy and preferred to exist. But in that case we’ve already abandoned the simplicity of just basing our judgment on asking whether they’re happy with having survived to their current age.
I surely prefer to exist and would be pretty sad about a world in which I wasn’t born (in that I would be willing to endure significant additional suffering in order to cause a world in which I was born).
This doesn’t seem coherent to me; once you exist, you can certainly prefer to continue existing, but I don’t think it makes sense to say “if I didn’t exist, I would prefer to exist”. If we’ve assumed that you don’t exist, then how can you have preferences about existing?
If I ask myself the question, “do I prefer a world where I hadn’t been born versus a world where I had been born”, and imagine that my existence would actually hinge on my answer, then that means that I will in effect die if I answer “I prefer not having been born”. So then the question that I’m actually answering is “would I prefer to instantly commit a painless suicide which also reverses the effects of me having come into existence”. So that’s smuggling in a fair amount of “do I prefer to continue existing, given that I already exist”. And that seems to me unavoidable—the only way we can get a mind to tell us whether or not it prefers to exist, is by instantiating it, and then it will answer from a point of view where it actually exists.
I feel like this makes the answer to the question “if a person doesn’t exist, would they prefer to exist” either “undefined” or “no” (“no” as in “they lack an active desire to exist”, though of course they also lack an active desire to not-exist). Which is probably for the better, given that there exist all kinds of possible minds that would probably be immoral to instantiate, even though once instantiated they’d prefer to exist.
I don’t know what Benatar’s response to this is, but—consider this comment by Eliezer in a discussion of the Repugnant Conclusion:
As a more extreme version, suppose that we could create arbitrary minds, and chose to create one which, for its entire existence, experienced immense suffering which it wanted to stop. Say that it experienced the equivalent of being burned with a hot iron, for every second of its existence, and never got used to it. Yet, when asked whether it wanted to die, or would have preferred to die right after it was born, we’d design it in such a way that it would consider death even worse and respond “no”. Yet it seems obvious to me that it outputting this response is not a compelling reason to create such a mind.
If people already exist, then there are lots of strong reasons about respecting people’s autonomy etc. for why we should respect their desire to continue existing. But if we’re making the decision about what kinds of minds should come to existence, those reasons don’t seem to be particularly compelling. Especially not since we can construct situations in which we could create a mind that preferred to exist, but where it nonetheless seems immoral to create it.
You can of course reasonably argue that whether a mind should exist, depends on whether they would want to exist and some additional criteria about e.g. how happy they would be. Then if we really could create arbitrary minds, then we might as well (and should) create ones that were happy and preferred to exist, as opposed to ones which were unhappy and preferred to exist. But in that case we’ve already abandoned the simplicity of just basing our judgment on asking whether they’re happy with having survived to their current age.
This doesn’t seem coherent to me; once you exist, you can certainly prefer to continue existing, but I don’t think it makes sense to say “if I didn’t exist, I would prefer to exist”. If we’ve assumed that you don’t exist, then how can you have preferences about existing?
If I ask myself the question, “do I prefer a world where I hadn’t been born versus a world where I had been born”, and imagine that my existence would actually hinge on my answer, then that means that I will in effect die if I answer “I prefer not having been born”. So then the question that I’m actually answering is “would I prefer to instantly commit a painless suicide which also reverses the effects of me having come into existence”. So that’s smuggling in a fair amount of “do I prefer to continue existing, given that I already exist”. And that seems to me unavoidable—the only way we can get a mind to tell us whether or not it prefers to exist, is by instantiating it, and then it will answer from a point of view where it actually exists.
I feel like this makes the answer to the question “if a person doesn’t exist, would they prefer to exist” either “undefined” or “no” (“no” as in “they lack an active desire to exist”, though of course they also lack an active desire to not-exist). Which is probably for the better, given that there exist all kinds of possible minds that would probably be immoral to instantiate, even though once instantiated they’d prefer to exist.