“Art of Rationality” is an oxymoron. Art follows (subjective) aesthetic principles; rationality follows (objective) evidence.
Science follows objective evidence. You’re not allowed to publish a paper where you conclude something based on a hunch, because anyone can claim they have a hunch. You can only do science with evidence that is undeniable. Not undeniably strong. You only need p = 0.05. But it has to be unquestionable that there really are those 4.3 bits of evidence.
Rationality follows subjective evidence. There often simply isn’t enough objective evidence to make all of your decisions. You have to use every trick in the book to make sure your model of reality is as accurate as possible.
It’s not about aesthetics, but this is using “art” in the sense of “more an art then a science,” which doesn’t imply aesthetics.
I would equate rationality with logic. Thus, the (subjective) priors are an input to rationality. LW Rationality appears to mix in a few subjective priors with the rationality.
That’s not what the word usually means on this site. You seem to be simultaneously objecting that (a) your idea of rationality is not optimal, and (b) LW rationality doesn’t perfectly follow your idea of rationality.
I wasn’t talking about priors. If you have a hunch because something is simpler, then that would be priors, but if you have a hunch because you’ve been subconsciously collecting evidence too vague to be put into words, then reality is causing the hunch, so it’s just evidence.
Science follows objective evidence. You’re not allowed to publish a paper where you conclude something based on a hunch, because anyone can claim they have a hunch. You can only do science with evidence that is undeniable. Not undeniably strong. You only need p = 0.05. But it has to be unquestionable that there really are those 4.3 bits of evidence.
Rationality follows subjective evidence. There often simply isn’t enough objective evidence to make all of your decisions. You have to use every trick in the book to make sure your model of reality is as accurate as possible.
It’s not about aesthetics, but this is using “art” in the sense of “more an art then a science,” which doesn’t imply aesthetics.
I would equate rationality with logic. Thus, the (subjective) priors are an input to rationality. LW Rationality appears to mix in a few subjective priors with the rationality.
That’s not what the word usually means on this site. You seem to be simultaneously objecting that (a) your idea of rationality is not optimal, and (b) LW rationality doesn’t perfectly follow your idea of rationality.
I wasn’t talking about priors. If you have a hunch because something is simpler, then that would be priors, but if you have a hunch because you’ve been subconsciously collecting evidence too vague to be put into words, then reality is causing the hunch, so it’s just evidence.