It’s funny that I push on the problem of moral nihilism just a little, and suddenly someone thinks I don’t believe in reality. :)
I’ve read the beginning and the end of the meta-ethics sequence, but not the middle. I agree with Eliezer that recursive questions are always possible, but you must stop asking them at some point or you miss more interesting issues. And I agree with his conclusion that the best formulation of modern ethics is consideration for the happiness of beings capable of recursive thought.
I like to write a discussion post (or a series of posts) on this issue, but I don’t know where to start. Someone else responded to me (EDIT: with what seemed to me like] questioning the assertion that science is a one-way ratchet, always getting better, never getting worse. [EDIT: But we don’t seem to have actually communicated at all, which isn’t a success on my part.]
In case you want a connection to Artificial Intelligence:
Eliezer talks about the importance of provably Friendly AI, and I agree with his point. If we create super-intelligence and it doesn’t care about our desires, that would be very bad for us. But I think that the problem I’m highlighting says something about the possibility of proving that an AI is Friendly.
Someone else responded to me by questioning the assertion that science is a one-way ratchet, always getting better, never getting worse.
It seems likely to me that I’m the person you’re referring to. If so, I don’t endorse your summary. More generally, I’m not sure either of us understood the other one clearly enough in that exchange to merit confident statements on either of our parts about what was actually said, short of literal quotes .
It’s funny that I push on the problem of moral nihilism just a little, and suddenly someone thinks I don’t believe in reality. :)
I’ve read the beginning and the end of the meta-ethics sequence, but not the middle. I agree with Eliezer that recursive questions are always possible, but you must stop asking them at some point or you miss more interesting issues. And I agree with his conclusion that the best formulation of modern ethics is consideration for the happiness of beings capable of recursive thought.
I like to write a discussion post (or a series of posts) on this issue, but I don’t know where to start. Someone else responded to me (EDIT: with what seemed to me like] questioning the assertion that science is a one-way ratchet, always getting better, never getting worse. [EDIT: But we don’t seem to have actually communicated at all, which isn’t a success on my part.]
In case you want a connection to Artificial Intelligence:
Eliezer talks about the importance of provably Friendly AI, and I agree with his point. If we create super-intelligence and it doesn’t care about our desires, that would be very bad for us. But I think that the problem I’m highlighting says something about the possibility of proving that an AI is Friendly.
It seems likely to me that I’m the person you’re referring to. If so, I don’t endorse your summary. More generally, I’m not sure either of us understood the other one clearly enough in that exchange to merit confident statements on either of our parts about what was actually said, short of literal quotes .