> If I knew I was above average at something then I could use it. But I'm using my faculties to think I'm above average and they might be wrong.
I find it fascinating the extent to which this mindset has permeated the rationality/effective altruism communities (despite admonitions from people like Eliezer). It's just not that bad if you sometimes think you're above average when you're not! As long as you course-correct quickly you'll be fine.
I think the problem here is that most people's typical thought processes are broken in *multiple* major ways:
1. We are tremendously overconfident and update too slowly on new evidence;
2. We are pathologically afraid of doing things that go badly, so require way too high of a bar of confidence to act on.
It's easy to demolish 1 for yourself by reading the right blogs; it's much harder to demolish 2 because it requires actually going out and doing things (while getting advice from people who are good at avoiding 2).
Amusingly, 1 and 2 sorta roughly cancel out at the 30,000-foot level, so demolishing only one is actually maladaptive. The result is a bunch of people who are great at noticing people being dumb, but in a way that is actually neutral/negative for their ability to go out and do things.
I agree broadly. The problem is a logical inconsistency that's theoretically true but practically probably immaterial ... I put it there for completeness though as in the other paras I'm against using that as a crutch to not create a winning strategy ...
> If I knew I was above average at something then I could use it. But I'm using my faculties to think I'm above average and they might be wrong.
I find it fascinating the extent to which this mindset has permeated the rationality/effective altruism communities (despite admonitions from people like Eliezer). It's just not that bad if you sometimes think you're above average when you're not! As long as you course-correct quickly you'll be fine.
I think the problem here is that most people's typical thought processes are broken in *multiple* major ways:
1. We are tremendously overconfident and update too slowly on new evidence;
2. We are pathologically afraid of doing things that go badly, so require way too high of a bar of confidence to act on.
It's easy to demolish 1 for yourself by reading the right blogs; it's much harder to demolish 2 because it requires actually going out and doing things (while getting advice from people who are good at avoiding 2).
Amusingly, 1 and 2 sorta roughly cancel out at the 30,000-foot level, so demolishing only one is actually maladaptive. The result is a bunch of people who are great at noticing people being dumb, but in a way that is actually neutral/negative for their ability to go out and do things.
I agree broadly. The problem is a logical inconsistency that's theoretically true but practically probably immaterial ... I put it there for completeness though as in the other paras I'm against using that as a crutch to not create a winning strategy ...