To be honest, Eliezer made a slightly different argument: 1) humans share (because of evolution) a psychological unity that is not affected by regional or temporal distinctions; 2) this unity entails a set of values that is inescapable for every human beings, its collective effect on human cognition and actions we dub “morality”; 3) Clippy, Elves and Pebblesorters, being fundamentally different, share a different set of values that guide their actions and what they care about; 4) those are perfectly coherent and sound for those who entertain them, we should though do not call them “Clippy’s, Elves’ or Pebblesorters’ morality”, because words should be used in such a way to maximize their usefulness in carving reality: since we cannot go out of our programming and conceivably find ourselves motivated by eggnog or primality, we should not use the term and instead use primality or other words. That’s it: you can debate any single point, but I think the difference is only formal. The underlying understanding, that “motivating set of values” is a two place predicate, is the same, Yudkowski preferred though to use different words for different partially applied predicates, on the grounds of point 1 and 4.
those are perfectly coherent and sound for those who entertain them, we should though do not call them “Clippy’s, Elves’ or Pebblesorters’ morality”, because words should be used in such a way to maximize their usefulness in carving reality: since we cannot go out of our programming and conceivably find ourselves motivated by eggnog or primality, we should not use the term and instead use primality or other words.
So my car is a car becuse it motor-vates me, but your car is no car at all, because it motor-vates you around, but not me. And yo mama ain’t no Mama cause she ain’t my Mama!
Yudkowsky isn’t being rigourous, he is instead appealing to an imaginary rule, one that is not seen in any other case.
And it’s not like the issue isn’t important, either .. obviously the premissibility of imposing ones values on others depends on whether they are immoral, amoral, differently moral , etc. Differrently moral is still a possibilirt, for the reasons that you are differently mothered, not unmohtered.
So my car is a car becuse it motor-vates me, but your car is no car at all, because it motor-vates you around, but not me.
The difference is not between two cars, yours and mine, but between a passegner ship and a cargo ship, built for two different purpose and two different class of users.
Yudkowsky isn’t being rigourous, he is instead appealing to an imaginary rule, one that is not seen in any other case.
On this we surely agree, I just find the new rule better than the old one. But this is the least important part of the whole discussion.
obviously the premissibility of imposing ones values on others depends on whether they are immoral, amoral, differently moral , etc. Differrently moral is still a possibilirt, for the reasons that you are differently mothered, not unmohtered.
This is well explored in “Three worlds collide”. Yudkowski vision of morality is such that it assigns different morality to different aliens, and the same morality to the same species (I’m using your convention). When different worlds collide, it is moral for us to stop babyeaters from eating babies, and it is moral for the superhappy to happify us. I think Eliezer is correct in showing that the only solution is avoiding contact at all.
The difference is not between two cars, yours and mine, but between a passegner ship and a cargo ship, built for two different purpose and two different class of users.
That seems different to what you were saying before.
This is well explored in “Three worlds collide”. Yudkowski vision of morality is such that it assigns different morality to different aliens, and the same morality to the same species (I’m using your convention). When different worlds collide, it is moral for us to stop babyeaters from eating babies, and it is moral for the superhappy to happify us. I think Eliezer is correct in showing that the only solution is avoiding contact at all.
There’s not much objectivity in that.
Why is it so important that our morality is the one that motivates us? People keep repeating it as though its a great revelation, but its equally true that babyeater morality motivates babyeaters, so the situation comes out looking symmetrical and therefore relativistc.
Maybe we should be abandoning the objectivity requirement as impossible. As I understand it this is in fact core to Yudkowsky’s theory- an “objective” morality would be the tablet he refers to as something to ignore.
I’m not entirely on Yudkowsky’s side in this. My view is that moral desires, whilst psychologically distinct from selfish desires, are not logically distinct and so the resolution to any ethical question is “What do I want?”. There is the prospect of coordination through shared moral wants, but there is the prospect of coordination through shared selfish wants as well. Ideas of “the good of society” or “objective ethical truth” are simply flawed concepts.
But I do think Yudkowsky has a good point both of you have been ignoring. His stone tablet analogy, if I remember correctly, sums it up.
“I think Eliezer is correct in showing that the only solution is avoiding contact at all.”: Assumes that there is such a thing as an objective solution, if implicitly.
“The difference is not between two cars, yours and mine, but between a passegner ship and a cargo ship, built for two different purpose and two different class of users.”: Passenger and cargo ships both have purposes within human morality. Alien moralities are likely to contradict each other.
“There’s not much objectivity in that.”: What if objectivity in the sense you describe is impossible?
“Why is it so important that our morality is the one that motivates us? People keep repeating it as though its a great revelation, but its equally true that babyeater morality motivates babyeaters, so the situation comes out looking symmetrical and therefore relativistc.”: If it isn’t, then it comes back to the amoralist challenge. Why should we even care?
Maybe we should be abandoning the objectivity requirement as impossible.
Maybe we should also consider in parallel the question of whether objectivity is necessary. If objectivity is both necessary to morality and impossible, then nihilism results.
The basic, pragmatic argument for the objectivity or quasi-objectivity of ethics is that it is connected to practices of reward and punishment, which either happen or not.
As I understand it this is in fact core to Yudkowsky’s theory- an “objective” morality would be the tablet he refers to as something to ignore.
I’m not entirely on Yudkowsky’s side in this. My view is that moral desires, whilst psychologically distinct from selfish desires, are not logically distinct and so the resolution to any ethical question is “What do I want?”.
if you are serious about the unselfish bit, then surely it boils down to “what do they want” or “what do we want”.
What if objectivity in the sense you describe is impossible?
i don’t accept the Moral Void argument, for the reasons given. Do you have another?
If it isn’t, then it comes back to the amoralist challenge. Why should we even care?
The idea that humans are uniquely motivated by human morality isn’t put forward as a an answer to the amoralist challenge, it is put forward as a a way of establishing something like moral objectivism.
“words should be used in such a way to maximize their usefulness in carving reality”
That does not mean that we should not use general words, but that we should have both general words and specific words. That is why it is right to speak of morality in general, and human morality in particular.
As I stated in other replies, it is not true that this disagreement is only about words. In general, when people disagree about how words should be used, that is because they disagree about what should be done. Because when you use words differently, you are likely to end up doing different things. And I gave concrete places where I disagree with Eliezer about what should be done, ways that correspond to how I disagree with him about morality.
In general I would describe the disagreement in the following way, although I agree that he would not accept this characterization: Eliezer believes that human values are intrinsically arbitrary. We just happen to value a certain set of things, and we might have happened to value some other random set. In whatever situation we found ourselves, we would have called those things “right,” and that would have been a name for the concrete values we had.
In contrast, I think that we value the things that are good for us. What is “good for us” is not arbitrary, but an objective fact about relationships between human nature and the world. Now there might well be other rational creatures and they might value other things. That will be because other things are good for them.
I agree that not everything in particular that people value is good for them. I say that everything that they value in a fundamental way is good for them. If you disagree, and think that some people value things that are bad for them in a fundamental way, how are they supposed to find out that those things are bad for them?
You are currently saying that the good is what people fundamentally value, and what people fundamentally value is good....for them. To escape vacuity, the second phrase would need to be cashed out as something like “side survival”.
But whose survival? If I fight for my tribe, I endanger my own survival, if I dodge the draft, I endanger my tribes’.
Real world ethics has a pretty clear answer: the group wins every time. Bravery beats cowardice, generosity beats meanness...these are human universals. if you reverse engineer that observation back into a theoretical understanding, you get the idea that morality is something programned into individuals by communities to promote the survival and thriving of communities.
But that is a rather different claim to The Good is the Good.
Clarification please. How do you avoid this supposed vacuity applying to basically all definitions? Taking a quick definition from a Google Search:
A: “I define a cat as a small domesticated carnivorous mammal with soft fur, a short snout, and retractile claws.”
B: “Yes, but is that a cat?”
Which could eventually lead back to A saying that:
A: “Yes you’ve said all these things, but it basically comes back to the claim a cat is a cat.”
Definitions are at best a record of usage. Usage can be broadened to include social practices such as reward and punishment. And the jails are full of people who commit theft (selfishness) , rape (ditto), etc. And the medals and plaudits go to the brave (altruism), the generous (ditto), etc.
I’m not sure how you’re addressing what I said. What do you mean by escaping vacuity? I used “good for them” in that comment because you did, when you said that not everything people value is good for them. I agree with that, if you mean the particular values that people have, but not in regard to their fundamental values.
Saying that something is morally good means “doing this thing, after considering all the factors, is good for me,” and saying that it is morally bad means “doing this thing, after considering all the factors, is bad for me.” Of course something might be somewhat good, without being morally good, because it is good according to some factors, but not after considering all of them. And of course whether or not it will benefit your communities is one of the factors.
I’m going to assume you mean what you say and are not just arguing about definitions. In that case:
You would be an apologist for HP Lovecraft’s Azathoth, at best, if you lived in his universe. There’s no objective criterion you could give to explain why that wouldn’t be moral, unless you beg the question and bring in moral criteria to judge a possible ‘ground of morality.’ Yes, I’m saying Nyarlathotep should follow morality instead of the supposed dictates of his alien god. And that’s not a contradiction but a tautology.
While I’m on the subject, Aquinian theology is an ugly vulgarization of Aristotle’s, the latter being more naturally linked to HPL’s Azathoth or the divine pirates of Pastafarianism.
That is why it is right to speak of morality in general, and human morality in particular.
I prefer Eliezer’s way because it makes evident, when talking to someone who hasn’t read the Sequence, that there are different set of self-consistent values, but it’s an agreement that people should have before starting to debate and I personally would have no problem in talking about different moralities.
Eliezer believes that human values are intrinsically arbitrary
But does he? Because that would be demonstrably false. Maybe arbitrary in the sense of “occupying a tiny space in the whole set of all possible values”, but since our morality is shaped by evolution, it will contain surely some historical accident but also a lot of useful heuristics. No human can value drinking poison, for example.
What is “good for us” is not arbitrary, but an objective fact about relationships between human nature and the world
If you were to unpack “good”, would you insert other meanings besides “what helps our survival”?
“There are different sets of self-consistent values.” This is true, but I do not agree that all logically possible sets of self-consistent values represent moralities. For example, it would be logically possible for an animal to value nothing but killing itself; but this does not represent a morality, because such an animal cannot exist in reality in a stable manner. It cannot come into existence in a natural way (namely by evolution) at all, even if you might be able to produce one artificially. If you do produce one artificially, it will just kill itself and then it will not exist.
This is part of what I was saying about how when people use words differently they hope to accomplish different things. I speak of morality in general, not to mean “logically consistent set of values”, but a set that could reasonably exist in the real word with a real intelligent being. In other words, restricting morality to human values is an indirect way of promoting the position that human values are arbitrary.
As I said, I don’t think Eliezer would accept that characterization of his position, and you give one reason why he would not. But he has a more general view where only some sets of values are possible for merely accidental reasons, namely because it just happens that things cannot evolve in other ways. I would say the contrary—it is not an accident that the value of killing yourself cannot evolve, but this is because killing yourself is bad.
And this kind of explains how “good” has to be unpacked. Good would be what tends to cause tendencies towards itself. Survival is one example, but not the only one, even if everything else will at least have to be consistent with that value. So e.g. not only is survival valued by intelligent creatures in all realistic conditions, but so is knowledge. So knowledge and survival are both good for all intelligent creatures. But since different creatures will produce their knowledge and survival in different ways, different things will be good for them in relation to these ends.
To be honest, Eliezer made a slightly different argument:
1) humans share (because of evolution) a psychological unity that is not affected by regional or temporal distinctions;
2) this unity entails a set of values that is inescapable for every human beings, its collective effect on human cognition and actions we dub “morality”;
3) Clippy, Elves and Pebblesorters, being fundamentally different, share a different set of values that guide their actions and what they care about;
4) those are perfectly coherent and sound for those who entertain them, we should though do not call them “Clippy’s, Elves’ or Pebblesorters’ morality”, because words should be used in such a way to maximize their usefulness in carving reality: since we cannot go out of our programming and conceivably find ourselves motivated by eggnog or primality, we should not use the term and instead use primality or other words.
That’s it: you can debate any single point, but I think the difference is only formal. The underlying understanding, that “motivating set of values” is a two place predicate, is the same, Yudkowski preferred though to use different words for different partially applied predicates, on the grounds of point 1 and 4.
So my car is a car becuse it motor-vates me, but your car is no car at all, because it motor-vates you around, but not me. And yo mama ain’t no Mama cause she ain’t my Mama!
Yudkowsky isn’t being rigourous, he is instead appealing to an imaginary rule, one that is not seen in any other case.
And it’s not like the issue isn’t important, either .. obviously the premissibility of imposing ones values on others depends on whether they are immoral, amoral, differently moral , etc. Differrently moral is still a possibilirt, for the reasons that you are differently mothered, not unmohtered.
The difference is not between two cars, yours and mine, but between a passegner ship and a cargo ship, built for two different purpose and two different class of users.
On this we surely agree, I just find the new rule better than the old one. But this is the least important part of the whole discussion.
This is well explored in “Three worlds collide”. Yudkowski vision of morality is such that it assigns different morality to different aliens, and the same morality to the same species (I’m using your convention). When different worlds collide, it is moral for us to stop babyeaters from eating babies, and it is moral for the superhappy to happify us. I think Eliezer is correct in showing that the only solution is avoiding contact at all.
That seems different to what you were saying before.
There’s not much objectivity in that.
Why is it so important that our morality is the one that motivates us? People keep repeating it as though its a great revelation, but its equally true that babyeater morality motivates babyeaters, so the situation comes out looking symmetrical and therefore relativistc.
Maybe we should be abandoning the objectivity requirement as impossible. As I understand it this is in fact core to Yudkowsky’s theory- an “objective” morality would be the tablet he refers to as something to ignore.
I’m not entirely on Yudkowsky’s side in this. My view is that moral desires, whilst psychologically distinct from selfish desires, are not logically distinct and so the resolution to any ethical question is “What do I want?”. There is the prospect of coordination through shared moral wants, but there is the prospect of coordination through shared selfish wants as well. Ideas of “the good of society” or “objective ethical truth” are simply flawed concepts.
But I do think Yudkowsky has a good point both of you have been ignoring. His stone tablet analogy, if I remember correctly, sums it up.
“I think Eliezer is correct in showing that the only solution is avoiding contact at all.”: Assumes that there is such a thing as an objective solution, if implicitly.
“The difference is not between two cars, yours and mine, but between a passegner ship and a cargo ship, built for two different purpose and two different class of users.”: Passenger and cargo ships both have purposes within human morality. Alien moralities are likely to contradict each other.
“There’s not much objectivity in that.”: What if objectivity in the sense you describe is impossible?
“Why is it so important that our morality is the one that motivates us? People keep repeating it as though its a great revelation, but its equally true that babyeater morality motivates babyeaters, so the situation comes out looking symmetrical and therefore relativistc.”: If it isn’t, then it comes back to the amoralist challenge. Why should we even care?
Maybe we should also consider in parallel the question of whether objectivity is necessary. If objectivity is both necessary to morality and impossible, then nihilism results.
The basic, pragmatic argument for the objectivity or quasi-objectivity of ethics is that it is connected to practices of reward and punishment, which either happen or not.
The essential problem with the tablet is that it offers conclusions as a fait accompli, with no justification of argument. The point does not generalise against objectivity morality.
if you are serious about the unselfish bit, then surely it boils down to “what do they want” or “what do we want”.
i don’t accept the Moral Void argument, for the reasons given. Do you have another?
The idea that humans are uniquely motivated by human morality isn’t put forward as a an answer to the amoralist challenge, it is put forward as a a way of establishing something like moral objectivism.
“words should be used in such a way to maximize their usefulness in carving reality”
That does not mean that we should not use general words, but that we should have both general words and specific words. That is why it is right to speak of morality in general, and human morality in particular.
As I stated in other replies, it is not true that this disagreement is only about words. In general, when people disagree about how words should be used, that is because they disagree about what should be done. Because when you use words differently, you are likely to end up doing different things. And I gave concrete places where I disagree with Eliezer about what should be done, ways that correspond to how I disagree with him about morality.
In general I would describe the disagreement in the following way, although I agree that he would not accept this characterization: Eliezer believes that human values are intrinsically arbitrary. We just happen to value a certain set of things, and we might have happened to value some other random set. In whatever situation we found ourselves, we would have called those things “right,” and that would have been a name for the concrete values we had.
In contrast, I think that we value the things that are good for us. What is “good for us” is not arbitrary, but an objective fact about relationships between human nature and the world. Now there might well be other rational creatures and they might value other things. That will be because other things are good for them.
But not everything people value is actually good for them. You are retaining the problem of equating morality with values.
I agree that not everything in particular that people value is good for them. I say that everything that they value in a fundamental way is good for them. If you disagree, and think that some people value things that are bad for them in a fundamental way, how are they supposed to find out that those things are bad for them?
You are currently saying that the good is what people fundamentally value, and what people fundamentally value is good....for them. To escape vacuity, the second phrase would need to be cashed out as something like “side survival”.
But whose survival? If I fight for my tribe, I endanger my own survival, if I dodge the draft, I endanger my tribes’.
Real world ethics has a pretty clear answer: the group wins every time. Bravery beats cowardice, generosity beats meanness...these are human universals. if you reverse engineer that observation back into a theoretical understanding, you get the idea that morality is something programned into individuals by communities to promote the survival and thriving of communities.
But that is a rather different claim to The Good is the Good.
Clarification please. How do you avoid this supposed vacuity applying to basically all definitions? Taking a quick definition from a Google Search: A: “I define a cat as a small domesticated carnivorous mammal with soft fur, a short snout, and retractile claws.” B: “Yes, but is that a cat?”
Which could eventually lead back to A saying that:
A: “Yes you’ve said all these things, but it basically comes back to the claim a cat is a cat.”
Definitions are at best a record of usage. Usage can be broadened to include social practices such as reward and punishment. And the jails are full of people who commit theft (selfishness) , rape (ditto), etc. And the medals and plaudits go to the brave (altruism), the generous (ditto), etc.
I’m not sure how you’re addressing what I said. What do you mean by escaping vacuity? I used “good for them” in that comment because you did, when you said that not everything people value is good for them. I agree with that, if you mean the particular values that people have, but not in regard to their fundamental values.
Saying that something is morally good means “doing this thing, after considering all the factors, is good for me,” and saying that it is morally bad means “doing this thing, after considering all the factors, is bad for me.” Of course something might be somewhat good, without being morally good, because it is good according to some factors, but not after considering all of them. And of course whether or not it will benefit your communities is one of the factors.
I’m going to assume you mean what you say and are not just arguing about definitions. In that case:
You would be an apologist for HP Lovecraft’s Azathoth, at best, if you lived in his universe. There’s no objective criterion you could give to explain why that wouldn’t be moral, unless you beg the question and bring in moral criteria to judge a possible ‘ground of morality.’ Yes, I’m saying Nyarlathotep should follow morality instead of the supposed dictates of his alien god. And that’s not a contradiction but a tautology.
While I’m on the subject, Aquinian theology is an ugly vulgarization of Aristotle’s, the latter being more naturally linked to HPL’s Azathoth or the divine pirates of Pastafarianism.
I’m pretty sure this is not an attempt at discussion, but an attempt to be insulting, so I won’t discuss it.
I prefer Eliezer’s way because it makes evident, when talking to someone who hasn’t read the Sequence, that there are different set of self-consistent values, but it’s an agreement that people should have before starting to debate and I personally would have no problem in talking about different moralities.
But does he? Because that would be demonstrably false. Maybe arbitrary in the sense of “occupying a tiny space in the whole set of all possible values”, but since our morality is shaped by evolution, it will contain surely some historical accident but also a lot of useful heuristics.
No human can value drinking poison, for example.
If you were to unpack “good”, would you insert other meanings besides “what helps our survival”?
“There are different sets of self-consistent values.” This is true, but I do not agree that all logically possible sets of self-consistent values represent moralities. For example, it would be logically possible for an animal to value nothing but killing itself; but this does not represent a morality, because such an animal cannot exist in reality in a stable manner. It cannot come into existence in a natural way (namely by evolution) at all, even if you might be able to produce one artificially. If you do produce one artificially, it will just kill itself and then it will not exist.
This is part of what I was saying about how when people use words differently they hope to accomplish different things. I speak of morality in general, not to mean “logically consistent set of values”, but a set that could reasonably exist in the real word with a real intelligent being. In other words, restricting morality to human values is an indirect way of promoting the position that human values are arbitrary.
As I said, I don’t think Eliezer would accept that characterization of his position, and you give one reason why he would not. But he has a more general view where only some sets of values are possible for merely accidental reasons, namely because it just happens that things cannot evolve in other ways. I would say the contrary—it is not an accident that the value of killing yourself cannot evolve, but this is because killing yourself is bad.
And this kind of explains how “good” has to be unpacked. Good would be what tends to cause tendencies towards itself. Survival is one example, but not the only one, even if everything else will at least have to be consistent with that value. So e.g. not only is survival valued by intelligent creatures in all realistic conditions, but so is knowledge. So knowledge and survival are both good for all intelligent creatures. But since different creatures will produce their knowledge and survival in different ways, different things will be good for them in relation to these ends.
Any virulently self-reproducing meme would be another.
This would be a long discussion, but there’s some truth in that, and some falsehood.