First of all, you should mention timeless decision theory, or at least superrationality. These concepts are useful for explaining why people’s intuition that they should not steal is not horribly misguided even if the thief cares about himself more and/or would needs it more than the previous owner. You touched on this by pointing out that the economy would collapse if everyone stole all the time, but I would suggest being more explicit.
Very strongly disagree, and not just because I’m sceptical about both. The article is supposed about consequentialism, not Yvain’s particular moral system. It should explain why you should apply your moral analysis to certain data (state of the world) instead of others (“rights”), but it shouldn’t get involved in how your moral analysis exactly works.
Yvain correctly mentions that you can be a paperclip maximiser and still be a perfect consequentialist.
UDT and TDT are decision theories, not “moral systems”. To the extent that consequentialism necessarily relies on some kind of decision theory—as is clearly the case, since it advocates choosing the optimal actions to take based on their outcomes—a brief mention of CDT, UDT and TDT explaining their relevance to consequentialist ethics (see e.g. the issue of “rule utilitarianism” vs. “action utilitarianism”) would have been appropriate.
I deleted a moderate wall of text because I think I understand what you mean now. I agree that two consequentialists sharing the same moral/utility function, but adopting different decision theories, will have to make different choices.
However, I don’t think it would be a very good idea to talk about various DTs in the FAQ. That is: showing that “people’s intuition that they should not steal is not horribly misguided”, by offering them the option of a DT that supports a similar rule, doesn’t seem to me like a worthy goal for the document. IMO, people should embrace consequentialism because it makes sense—because it doesn’t rely on pies in the sky—not because it can be made to match their moral intuitions. If you use that approach, you could in the same way use the fat man trolley problem to support deontology.
I might be misinterpreting you or taking this too far, but what you suggest sounds to me like “Let’s write ‘Theft is wrong’ on the bottom line because that’s what is expected by readers and makes them comfortable, then let’s find a consequentialist process that will give that result so they will be happy” (note that it’s irrelevant whether that process happens to be correct or wrong). I think discouraging that type of reasoning is even more important than promoting consequentialism.
people should embrace consequentialism because it makes sense—because it doesn’t rely on pies in the sky—not because it can be made to match their moral intuitions.
The whole point of CEV, reflexive consistency and the meta-ethics sequence is that morality is based on our intuitions.
Yes, I personally think that’s awful. LessWrong rightly tendstopromote being sceptical of one’s mere intuitions in most contexts, and I think the same approach should be taken with morality (basically, this post on steroids).
Very strongly disagree, and not just because I’m sceptical about both. The article is supposed about consequentialism, not Yvain’s particular moral system. It should explain why you should apply your moral analysis to certain data (state of the world) instead of others (“rights”), but it shouldn’t get involved in how your moral analysis exactly works.
Yvain correctly mentions that you can be a paperclip maximiser and still be a perfect consequentialist.
UDT and TDT are decision theories, not “moral systems”. To the extent that consequentialism necessarily relies on some kind of decision theory—as is clearly the case, since it advocates choosing the optimal actions to take based on their outcomes—a brief mention of CDT, UDT and TDT explaining their relevance to consequentialist ethics (see e.g. the issue of “rule utilitarianism” vs. “action utilitarianism”) would have been appropriate.
I deleted a moderate wall of text because I think I understand what you mean now. I agree that two consequentialists sharing the same moral/utility function, but adopting different decision theories, will have to make different choices.
However, I don’t think it would be a very good idea to talk about various DTs in the FAQ. That is: showing that “people’s intuition that they should not steal is not horribly misguided”, by offering them the option of a DT that supports a similar rule, doesn’t seem to me like a worthy goal for the document. IMO, people should embrace consequentialism because it makes sense—because it doesn’t rely on pies in the sky—not because it can be made to match their moral intuitions. If you use that approach, you could in the same way use the fat man trolley problem to support deontology.
I might be misinterpreting you or taking this too far, but what you suggest sounds to me like “Let’s write ‘Theft is wrong’ on the bottom line because that’s what is expected by readers and makes them comfortable, then let’s find a consequentialist process that will give that result so they will be happy” (note that it’s irrelevant whether that process happens to be correct or wrong). I think discouraging that type of reasoning is even more important than promoting consequentialism.
The whole point of CEV, reflexive consistency and the meta-ethics sequence is that morality is based on our intuitions.
Yes, I personally think that’s awful. LessWrong rightly tends to promote being sceptical of one’s mere intuitions in most contexts, and I think the same approach should be taken with morality (basically, this post on steroids).
If this is to be useful, it would have to read “that our intuitions are based on morality”.