I found the bit at the end contrasting Bayesian truth-seeking posture from believing-in and fighting-to-change-something posture was helpful. Interesting that you’ve seen lots of LessWrongers just passively watch to see if something will go well instead of fighting for it because of the Bayesian posture.
It seems to me there’s an important analog to descriptive statistics versus causal inference. The probability of X given Y is actually completely different from the probability of X given that I make Y happen. You can convert between them in aggregate sometimes with the do-calculus, so there is a link, but the one can be very different from the other in any given case. A thing can have really low probability in the base rate, or “if I do nothing about it” and yet have much higher probability if we rally around it to effect a change.
I found the bit at the end contrasting Bayesian truth-seeking posture from believing-in and fighting-to-change-something posture was helpful. Interesting that you’ve seen lots of LessWrongers just passively watch to see if something will go well instead of fighting for it because of the Bayesian posture.
It seems to me there’s an important analog to descriptive statistics versus causal inference. The probability of X given Y is actually completely different from the probability of X given that I make Y happen. You can convert between them in aggregate sometimes with the do-calculus, so there is a link, but the one can be very different from the other in any given case. A thing can have really low probability in the base rate, or “if I do nothing about it” and yet have much higher probability if we rally around it to effect a change.