I was just rereading Three Worlds Collide today and noticed that my feelings about the ending have changed over the last few years. It used to be obvious to me that the “status quo” ending was better. Now I feel that the “super happy” ending is better, and it’s not just a matter of feelings—it’s somehow axiomatically better, based on what I know about decision theory.
Namely, the story says that the super happies are smarter and understand humanity’s utility function better, and also that they are moral and wouldn’t offer a deal unless it was beneficial according to both utility functions being merged (not just according to their value of happiness). Under these conditions, accepting the deal seems like the right thing to do.
Does the story actually says the Superhappies really know humanity’s utility function better? As in, does an omniscient narrator tell it, or is it a Superhappy or one of the crew that says this? That changes a lot, to me. Of course the Superhappies would believe they know our utility function better than we do. Just like how the humans assumed they knew what was better for the Babyeaters.
Similarly, the Superhappies are moral, for their idea of morality. They were perfectly willing to use force (not physical, but force nonetheless) to encourage humans to see their point of view. They threatened humanity and were willing to forcibly change human children, even if the adults could continue to feel pain. While humans also employs threats and force to change behavior, in most cases we would be hard-pressed to call that “moral.”
From a meta-perspective, I’d findit odd if Yudkowsky wrote it like that. He’s not careless enough to make that mistake and as far as I know, he thinks humanity’s utility function goes beyond mere bliss.
The only way I think you could see the Superhappies’ solution as acceptable if you don’t think jokes or fiction (or other sort of arts involving “deception”) are something humans would value as part of their utility function. Which I personally would find very hard to understand.
The only way I think you could see the Superhappies’ solution as acceptable if you don’t think jokes or fiction (or other sort of arts involving “deception”) are something humans would value as part of their utility function.
Um, that’s the opposite of how utility functions work. They don’t have sacred components. You can and should trade off one component for a larger gain in another component. That’s exactly what the super happies were offering.
What why would this be true? Utility functions don’t have to be linear, it could even be the case that I place no additional utility on happiness beyond a certain level.
True, but the question in the story is whether total cost of suffering > total benefit from being able to suffer. These are the components being traded. When put this way, the question answers itself. The only reason to reply “no” is status quo bias (mentioning sacred components of utility is an example of that). The standard fix for that is the reversal test: do you think the current amount of suffering is coincidentally exactly optimal, or would you prefer to add some more? That test is actually mentioned in the story, the humans apply it to babyeaters, but forget to apply it to themselves.
the question in the story is whether total cost of suffering > total benefit from being able to suffer
The answer to this question is “No.”
do you think the current amount of suffering is coincidentally exactly optimal, or would you prefer to add some more?
Some people could use more. Many others could use less.
The question you should ask first is whether being able to suffer is a good thing or a bad thing. You start with the assumption that it is bad, that suffering is bad. You do not sufficiently investigate what the alternative is; you do not sufficiently consider that experience is subjective, and subjectivity requires reference points. To eliminate, in perpetuity, that half of the axis below the current reference point, is to eliminate the axis entirely.
There’s no way of calculating utility, period. The issue is more substantively that suffering is relative, and that the elimination of suffering is also the elimination of happiness.
Happiness and suffering are the same thing—the experience of a divergence from the norm of your well-being, your ground state. They just differ in direction.
A long time ago, I experienced both. For most of my life, I experienced neither—you think pain is a negative experience, I found it to be an -interesting- experience, a diversion from the endless gray. Today, I experience… a very limited degree of both, as a result of gradually accepting that suffering is the cost paid to experience happiness.
Equanimity, as it transpires, isn’t something you can experience only with regard to those things you don’t want to directly experience.
True, the difference is the direction, but surely that counts for something? Pain and pleasure are chemically and neurologically different phenomena. A ground state of “endless gray” is not something you’d really want.
suffering is the cost paid to experience happiness
I’m guessing you may be a Roman Catholic. In case you’re not, how did you come to see suffering as having exchange value?
I hope my comments are not taken as offensive. I know I sometimes tend to dramatize my degree of surprise. I genuinely wish to understand your position.
True, the difference is the direction, but surely that counts for something? Pain and pleasure are chemically and neurologically different phenomena.
I still don’t experience “pleasure”, at least in sense where I can say, “Yes, that sensation is positive in a way other sensations are not”. At best I can say I experience variety. Pain is just starting to be a negative thing; it’s difficult to accept it as suffering when it was one of the few things that offered any variety at all to my experience for many years. Pain isn’t pleasure, they’re different flavors, but they’re both spices.
A ground state of “endless gray” is not something you’d really want.
This is very true.
I’m guessing you may be a Roman Catholic. In case you’re not, how did you come to see suffering as having exchange value?
I was raised, and remain, an atheist. And exchange value isn’t quite the same thing; it’s more they’re the same variable, but different values. Living for more than a decade without either suffering or happiness, and only starting to experience happiness when I started to allow myself to experience suffering.
I regard suffering and happiness as sums, rather than independent variables; they’re composite emotions, perhaps better modeled as waves, created by summing up one’s current total mindstate. Each is the inverse of the other; being waves, rather than simple linear values, it’s possible to both be suffering and be happy, if one area of one’s life is going well and one area is going poorly. But they’re both invariably tied to one’s norm; if one has had a consistently good life, their life continually to be consistently good isn’t going to provide any happiness, even though the same section of life, transplanted into somebody with a consistently bad life, would provide ecstasy. Likewise, a consistently bad life doesn’t translate into suffering; it’s the particularly bad parts of that life that are experienced as suffering, everything else is experienced as the norm.
This is backed up by studies of self-reported happiness, which tracks a norm, and only rarely [ETA: permanently] deviates from that norm. This norm, this base level of self-reported happiness (which I distinguish from experienced happiness), is the norm from which happiness and suffering are experienced as deviations.
this base level of self-reported happiness … is the norm from which happiness and suffering are experienced as deviations.
True, but only partially true. The stable base level, as you know, varies. There are people with high-happiness stable level and people with low-happiness stable level. These people look and behave very differently in real life. The high-base people look and behave happy at their neutral setting I don’t see any reason to believe that it’s just outwards manifestations which do not reflect the internal state. The low-base people are, in contrast, much less happy at their neutral setting.
So yes, on the one hand happiness/suffering is relative to your base state; but on the other hand there is an absolute scale as well and high-base people are happier than low-base people.
It’s hard to say what goes on in other people’s heads, but my self-reported happiness would be an assessment of my well-being relative to what I regard as my cultural norm, whereas my experienced happiness is a different value entirely.
I base my belief that this is the norm for humans on the fact that life satisfaction decreases are correlated with suicide rates irrespective of the absolute value of life satisfaction (although certain factors can have an inhibitive effect); that is, wealthy nations, which generally have higher self-reported happiness levels, also have high suicide rates. Their high base level of happiness, if this were the same variable as experienced happiness, should otherwise offset the suffering they experience, which does not appear to happen.
People’s social behavior is more predicated on their perceived relationship to the local/current social group than the state of their internal variables. I don’t base this on any study, but rather personal observation.
I’m not talking about evaluating one’s own internal state. I’m talking about outward signs.
I know both high-base and low-base people from, more or less, the same cultural circles. It’s not that they would answer the question “How happy are you?” differently—I don’t know, I haven’t asked. It’s just that the high-base people smile and laugh a lot, are prone to engaging in spontaneous fun, are generally comfortable with life. And the low-base people tend to have a characteristic disapproving expression on their faces (which will actually mold their face by middle age), whine and grumble a lot, and find life generally unpleasant.
Note that here I’m talking about, basically, long-term averages. In the short term high-base people can and will get unhappy and depressed; low-base people can and will get excited and joyful. But both will revert to the mean—I’m not talking about bipolar people who will oscillate between highs and lows, they are a separate category.
At the start, I decided that emotions were holding me back, and that logic was the more appropriate path, and so sat down one day and destroyed my emotions.
Over the next few years? I graduated high school, then college, got a couple of low-level jobs, then a real job, which I’ve held since. Dated a few people, role-played a normal person in the course of my interactions with them.
My emotions weren’t completely gone, over this period of time, but rather… remote, happening to somebody else. If they got particularly intense, I could observe my body’s reaction to them—hands clenching in anger, for example—but I didn’t actually experience them. The emotions were there, but the connection to my conscious mind was severed.
At some point in there, I read Atlas Shrugged, which convinced me that emotions were not, in fact, useless distractions from pure logic. I still wasn’t experiencing them, but the absence was no longer desirable; at that point, it was neutral. Everything was neutral, really. That began the gray phase of my life.
I honestly don’t remember much from that period of time. Nothing had any kind of significance. I worked, I dated, read books, played games. None of it particularly mattered; existence was a habit without importance. It wasn’t unpleasant, because unpleasantness would have been something. I was told I was depressed. If I was, if I wasn’t—didn’t particularly matter.
Then I tried LSD. And… I had a day that wasn’t gray. I appreciated the green color of the leaves on trees, the texture of the bark. Shrug So I decided I would prefer to live like that all the time, and started permitting myself to experience life again. Started taking Vitamin D, which kick-started the process.
Which began a rather dark period, as allowing myself to experience required confronting all the suffering I had avoided. The deaths of some people who had been close to me in my youth. An ex-girlfriend raping me, and before that with another partner, my first sexual encounters having been undesired, but my having not refused because I didn’t care enough to. How inherently abusive many of the relationships I was in were, how dysfunctional the situation I had allowed myself to get into was. Admitting to myself that much of the past decade of my life had been a failure.
And then things got better, because the recognition that things were bad was the same as the recognition that things could get better, and so I starting making things better. I got out of the situation, and have started working towards the next phase in my life.
That’s an amazing journey of self-discovery. I, too, had a period where I wanted to erase the parts of me that I found useless, but I didn’t go as deeply Vulcan as you did. (You’re the first person I’ve met who became more sensitive and overall nicer because of Atlas Shrugged.) I’m sorry to hear that you went through so many dark places during your process, and I find your final meditations on the meaning of suffering to be quite inspiring. You have my admiration.
I think what the “true” (status-quo) ending proves is that the Super-Happies did not accurately model humanity’s utility function at all. If they had, they would have proposed a deal where humanity gets rid of most of its pain, but still keeps some, especially those “grim” things that humans actually like (somewhat counter-intuitively). (And perhaps the Babyeaters’ thing would then be understood as one of these “grim” things by humans, as it clearly is for the Babyeaters themselves It’s not clear if the Superhappies would be willing to acquire this value, though). This is a deal that humans would indeed accept, since it agrees with their values. I think the true moral of this story is that getting human wants right for something like CEV is a hard problem, and making even small mistakes can have big consequences.
My feeling is that many utility functions in the general class of utility functions that the super happy’s is drawn from would lie about how advantageous it is to merge. Weren’t the humans going to lie to the babyeaters?
But it’s still a compromise. Is it part of humanity’s utility function to value another species’ utility function to such an extent that they would accept the tradeoff of changing humanity’s utility function to preserve as much of the other species’ utility function?
I don’t recall any mention of humanity being total utilitarians in the story. Neither did the compromise made by the superhappies strike me as being better for all parties than their original values were, for each of them.
The only reason the compromise was supposed to be beneficial is because the three species made contact and couldn’t easily coexist together from that point on. Also, because the superhappies were the stronger force and could therefore easily enforce their own solution. Cutting off the link removes those assumptions, and allows each species to preserve its utility function, which I assume they have a preference for, at least humans and baby-eaters.
Cutting off the link (...) allows each species to preserve its utility function, which I assume they have a preference for, at least humans and baby-eaters.
There was an asymetry in the story, if I remember correctly.
Babyeaters had a preference for other species eating their babies. Humans and superhappies had a preference for other species not eating their babies. This part was symetrical. Superhappies also had a preference for other species never feeling any pain. But humans didn’t have a preference for other species feeling pain; they just wanted to more or less preserve their own biological status quo. They didn’t mind if superhappies remain… superhappy.
This is why cutting the link harms the superhappy utility function more than the human utility function. -- Humans will feel the relief that babyeater children are still saved by superhappies, more quickly and reliably than humans could do. On the other hand, superhappies will know that somewhere in the universe human babies are feeling pain and frustration, and there is nothing the superhappies can do about it.
The asymetry was that superhappies didn’t seem ethically repulsive to humans. Well, apart from what they wanted to do with humans; which was successfully avoided.
In the story the superhappies propose to self-modify to appreciate complex art, not just simple porn, and they say that humans and babyeaters will both think that is an improvement. So to some degree the superhappies (with their very ugly spaceships) are repulsive to humans, although not as strongly repulsive as the babyeaters.
they are moral and wouldn’t offer a deal unless it was beneficial according to both utility functions being merged (not just according to their value of happiness).
I guess whether it is beneficial or not depends on what you compare to? They say,
he obvious starting point upon which to build further negotiations, is to combine and compromise the utility functions of the three species until we mutually satisfice, providing compensation for all changes demanded.
So they are aiming for satisficing rather than maximizing utility: according to all three before-the-change moralities, the post-change state of affairs should be acceptable, but not necessarily optimal. Consider these possibilities:
1) Baby-eaters are modified to no longer eat sentient babies; humans are unchanged; Superhappies like art.
2) Baby-eaters are modified to no longer eat sentient babies; humans are pain-free and eat babies; Superhappies like art.
3) Baby-eaters, humans, and Superhappies are all unchanged.
I think the intention of the author is that, according to pre-change human morality, (1) is the optimal choice, (2) is bad but acceptable, and (3) is unacceptable. The superhappies in the story claim that (2) is the only alternative that is acceptable to all three pre-change moralities. So the super-happy ending is beneficial in the sense that it avoids (3), but it’s a “bad” ending because it fails to get (1).
I was just rereading Three Worlds Collide today and noticed that my feelings about the ending have changed over the last few years. It used to be obvious to me that the “status quo” ending was better. Now I feel that the “super happy” ending is better, and it’s not just a matter of feelings—it’s somehow axiomatically better, based on what I know about decision theory.
Namely, the story says that the super happies are smarter and understand humanity’s utility function better, and also that they are moral and wouldn’t offer a deal unless it was beneficial according to both utility functions being merged (not just according to their value of happiness). Under these conditions, accepting the deal seems like the right thing to do.
Does the story actually says the Superhappies really know humanity’s utility function better? As in, does an omniscient narrator tell it, or is it a Superhappy or one of the crew that says this? That changes a lot, to me. Of course the Superhappies would believe they know our utility function better than we do. Just like how the humans assumed they knew what was better for the Babyeaters.
Similarly, the Superhappies are moral, for their idea of morality. They were perfectly willing to use force (not physical, but force nonetheless) to encourage humans to see their point of view. They threatened humanity and were willing to forcibly change human children, even if the adults could continue to feel pain. While humans also employs threats and force to change behavior, in most cases we would be hard-pressed to call that “moral.”
From a meta-perspective, I’d findit odd if Yudkowsky wrote it like that. He’s not careless enough to make that mistake and as far as I know, he thinks humanity’s utility function goes beyond mere bliss.
The only way I think you could see the Superhappies’ solution as acceptable if you don’t think jokes or fiction (or other sort of arts involving “deception”) are something humans would value as part of their utility function. Which I personally would find very hard to understand.
Um, that’s the opposite of how utility functions work. They don’t have sacred components. You can and should trade off one component for a larger gain in another component. That’s exactly what the super happies were offering.
What I’m saying is that humans aren’t wrong in trading off some amount of comfort so they can have jokes, fiction, art and romantic love.
What why would this be true? Utility functions don’t have to be linear, it could even be the case that I place no additional utility on happiness beyond a certain level.
True, but the question in the story is whether total cost of suffering > total benefit from being able to suffer. These are the components being traded. When put this way, the question answers itself. The only reason to reply “no” is status quo bias (mentioning sacred components of utility is an example of that). The standard fix for that is the reversal test: do you think the current amount of suffering is coincidentally exactly optimal, or would you prefer to add some more? That test is actually mentioned in the story, the humans apply it to babyeaters, but forget to apply it to themselves.
The answer to this question is “No.”
Some people could use more. Many others could use less.
The question you should ask first is whether being able to suffer is a good thing or a bad thing. You start with the assumption that it is bad, that suffering is bad. You do not sufficiently investigate what the alternative is; you do not sufficiently consider that experience is subjective, and subjectivity requires reference points. To eliminate, in perpetuity, that half of the axis below the current reference point, is to eliminate the axis entirely.
Do you have a proof for this? As far as I know, we have no universally agreed upon way to compare different ways of calculating utility.
There’s no way of calculating utility, period. The issue is more substantively that suffering is relative, and that the elimination of suffering is also the elimination of happiness.
Please explain in more detail. The Buddhist part of my brain just had a spit-take upon reading that.
Happiness and suffering are the same thing—the experience of a divergence from the norm of your well-being, your ground state. They just differ in direction.
A long time ago, I experienced both. For most of my life, I experienced neither—you think pain is a negative experience, I found it to be an -interesting- experience, a diversion from the endless gray. Today, I experience… a very limited degree of both, as a result of gradually accepting that suffering is the cost paid to experience happiness.
Equanimity, as it transpires, isn’t something you can experience only with regard to those things you don’t want to directly experience.
True, the difference is the direction, but surely that counts for something? Pain and pleasure are chemically and neurologically different phenomena. A ground state of “endless gray” is not something you’d really want.
I’m guessing you may be a Roman Catholic. In case you’re not, how did you come to see suffering as having exchange value?
I hope my comments are not taken as offensive. I know I sometimes tend to dramatize my degree of surprise. I genuinely wish to understand your position.
I still don’t experience “pleasure”, at least in sense where I can say, “Yes, that sensation is positive in a way other sensations are not”. At best I can say I experience variety. Pain is just starting to be a negative thing; it’s difficult to accept it as suffering when it was one of the few things that offered any variety at all to my experience for many years. Pain isn’t pleasure, they’re different flavors, but they’re both spices.
This is very true.
I was raised, and remain, an atheist. And exchange value isn’t quite the same thing; it’s more they’re the same variable, but different values. Living for more than a decade without either suffering or happiness, and only starting to experience happiness when I started to allow myself to experience suffering.
I regard suffering and happiness as sums, rather than independent variables; they’re composite emotions, perhaps better modeled as waves, created by summing up one’s current total mindstate. Each is the inverse of the other; being waves, rather than simple linear values, it’s possible to both be suffering and be happy, if one area of one’s life is going well and one area is going poorly. But they’re both invariably tied to one’s norm; if one has had a consistently good life, their life continually to be consistently good isn’t going to provide any happiness, even though the same section of life, transplanted into somebody with a consistently bad life, would provide ecstasy. Likewise, a consistently bad life doesn’t translate into suffering; it’s the particularly bad parts of that life that are experienced as suffering, everything else is experienced as the norm.
This is backed up by studies of self-reported happiness, which tracks a norm, and only rarely [ETA: permanently] deviates from that norm. This norm, this base level of self-reported happiness (which I distinguish from experienced happiness), is the norm from which happiness and suffering are experienced as deviations.
True, but only partially true. The stable base level, as you know, varies. There are people with high-happiness stable level and people with low-happiness stable level. These people look and behave very differently in real life. The high-base people look and behave happy at their neutral setting I don’t see any reason to believe that it’s just outwards manifestations which do not reflect the internal state. The low-base people are, in contrast, much less happy at their neutral setting.
So yes, on the one hand happiness/suffering is relative to your base state; but on the other hand there is an absolute scale as well and high-base people are happier than low-base people.
It’s hard to say what goes on in other people’s heads, but my self-reported happiness would be an assessment of my well-being relative to what I regard as my cultural norm, whereas my experienced happiness is a different value entirely.
I base my belief that this is the norm for humans on the fact that life satisfaction decreases are correlated with suicide rates irrespective of the absolute value of life satisfaction (although certain factors can have an inhibitive effect); that is, wealthy nations, which generally have higher self-reported happiness levels, also have high suicide rates. Their high base level of happiness, if this were the same variable as experienced happiness, should otherwise offset the suffering they experience, which does not appear to happen.
People’s social behavior is more predicated on their perceived relationship to the local/current social group than the state of their internal variables. I don’t base this on any study, but rather personal observation.
I’m not talking about evaluating one’s own internal state. I’m talking about outward signs.
I know both high-base and low-base people from, more or less, the same cultural circles. It’s not that they would answer the question “How happy are you?” differently—I don’t know, I haven’t asked. It’s just that the high-base people smile and laugh a lot, are prone to engaging in spontaneous fun, are generally comfortable with life. And the low-base people tend to have a characteristic disapproving expression on their faces (which will actually mold their face by middle age), whine and grumble a lot, and find life generally unpleasant.
Note that here I’m talking about, basically, long-term averages. In the short term high-base people can and will get unhappy and depressed; low-base people can and will get excited and joyful. But both will revert to the mean—I’m not talking about bipolar people who will oscillate between highs and lows, they are a separate category.
What happened to you during those years? Feel free to decline to answer if I’m being too intrusive.
At the start, I decided that emotions were holding me back, and that logic was the more appropriate path, and so sat down one day and destroyed my emotions.
Over the next few years? I graduated high school, then college, got a couple of low-level jobs, then a real job, which I’ve held since. Dated a few people, role-played a normal person in the course of my interactions with them.
My emotions weren’t completely gone, over this period of time, but rather… remote, happening to somebody else. If they got particularly intense, I could observe my body’s reaction to them—hands clenching in anger, for example—but I didn’t actually experience them. The emotions were there, but the connection to my conscious mind was severed.
At some point in there, I read Atlas Shrugged, which convinced me that emotions were not, in fact, useless distractions from pure logic. I still wasn’t experiencing them, but the absence was no longer desirable; at that point, it was neutral. Everything was neutral, really. That began the gray phase of my life.
I honestly don’t remember much from that period of time. Nothing had any kind of significance. I worked, I dated, read books, played games. None of it particularly mattered; existence was a habit without importance. It wasn’t unpleasant, because unpleasantness would have been something. I was told I was depressed. If I was, if I wasn’t—didn’t particularly matter.
Then I tried LSD. And… I had a day that wasn’t gray. I appreciated the green color of the leaves on trees, the texture of the bark. Shrug So I decided I would prefer to live like that all the time, and started permitting myself to experience life again. Started taking Vitamin D, which kick-started the process.
Which began a rather dark period, as allowing myself to experience required confronting all the suffering I had avoided. The deaths of some people who had been close to me in my youth. An ex-girlfriend raping me, and before that with another partner, my first sexual encounters having been undesired, but my having not refused because I didn’t care enough to. How inherently abusive many of the relationships I was in were, how dysfunctional the situation I had allowed myself to get into was. Admitting to myself that much of the past decade of my life had been a failure.
And then things got better, because the recognition that things were bad was the same as the recognition that things could get better, and so I starting making things better. I got out of the situation, and have started working towards the next phase in my life.
That’s an amazing journey of self-discovery. I, too, had a period where I wanted to erase the parts of me that I found useless, but I didn’t go as deeply Vulcan as you did. (You’re the first person I’ve met who became more sensitive and overall nicer because of Atlas Shrugged.) I’m sorry to hear that you went through so many dark places during your process, and I find your final meditations on the meaning of suffering to be quite inspiring. You have my admiration.
Pain and suffering are not the same thing. One woman will suffer while giving birth while the next doesn’t and enjoys the experience.
I think what the “true” (status-quo) ending proves is that the Super-Happies did not accurately model humanity’s utility function at all. If they had, they would have proposed a deal where humanity gets rid of most of its pain, but still keeps some, especially those “grim” things that humans actually like (somewhat counter-intuitively). (And perhaps the Babyeaters’ thing would then be understood as one of these “grim” things by humans, as it clearly is for the Babyeaters themselves It’s not clear if the Superhappies would be willing to acquire this value, though). This is a deal that humans would indeed accept, since it agrees with their values. I think the true moral of this story is that getting human wants right for something like CEV is a hard problem, and making even small mistakes can have big consequences.
My feeling is that many utility functions in the general class of utility functions that the super happy’s is drawn from would lie about how advantageous it is to merge. Weren’t the humans going to lie to the babyeaters?
But it’s still a compromise. Is it part of humanity’s utility function to value another species’ utility function to such an extent that they would accept the tradeoff of changing humanity’s utility function to preserve as much of the other species’ utility function?
I don’t recall any mention of humanity being total utilitarians in the story. Neither did the compromise made by the superhappies strike me as being better for all parties than their original values were, for each of them.
The only reason the compromise was supposed to be beneficial is because the three species made contact and couldn’t easily coexist together from that point on. Also, because the superhappies were the stronger force and could therefore easily enforce their own solution. Cutting off the link removes those assumptions, and allows each species to preserve its utility function, which I assume they have a preference for, at least humans and baby-eaters.
There was an asymetry in the story, if I remember correctly.
Babyeaters had a preference for other species eating their babies. Humans and superhappies had a preference for other species not eating their babies. This part was symetrical. Superhappies also had a preference for other species never feeling any pain. But humans didn’t have a preference for other species feeling pain; they just wanted to more or less preserve their own biological status quo. They didn’t mind if superhappies remain… superhappy.
This is why cutting the link harms the superhappy utility function more than the human utility function. -- Humans will feel the relief that babyeater children are still saved by superhappies, more quickly and reliably than humans could do. On the other hand, superhappies will know that somewhere in the universe human babies are feeling pain and frustration, and there is nothing the superhappies can do about it.
The asymetry was that superhappies didn’t seem ethically repulsive to humans. Well, apart from what they wanted to do with humans; which was successfully avoided.
In the story the superhappies propose to self-modify to appreciate complex art, not just simple porn, and they say that humans and babyeaters will both think that is an improvement. So to some degree the superhappies (with their very ugly spaceships) are repulsive to humans, although not as strongly repulsive as the babyeaters.
I guess whether it is beneficial or not depends on what you compare to? They say,
So they are aiming for satisficing rather than maximizing utility: according to all three before-the-change moralities, the post-change state of affairs should be acceptable, but not necessarily optimal. Consider these possibilities:
1) Baby-eaters are modified to no longer eat sentient babies; humans are unchanged; Superhappies like art.
2) Baby-eaters are modified to no longer eat sentient babies; humans are pain-free and eat babies; Superhappies like art.
3) Baby-eaters, humans, and Superhappies are all unchanged.
I think the intention of the author is that, according to pre-change human morality, (1) is the optimal choice, (2) is bad but acceptable, and (3) is unacceptable. The superhappies in the story claim that (2) is the only alternative that is acceptable to all three pre-change moralities. So the super-happy ending is beneficial in the sense that it avoids (3), but it’s a “bad” ending because it fails to get (1).
Hmm, I guess I interpreted the super happies proposal differently, as saying that humans get compensation for any downgrade from (1) to (2).