This has to be carefully targeted to a select group of people or the world would be drowning in noise (even more than it already is). I think many people tend to be overconfident in their own ideas and already ramble more and filter less than they ideally should. Fear of being wrong might help counteract that somewhat so I’m wary of trying to remove that counterweight in someone before making sure to fix the overconfidence problem, unless you’re only targeting the intervention on people who already aren’t overconfident.
I think both overconfidence and underconfidence are widespread, so it’s hard to tell which advice would do more good. Maybe we can agree that people tend to over-share their conclusions and under-share their evidence? That seems plausible overall; advising people to shift toward sharing evidence might help address both underconfidence (because evidence feels safer to share) and overconfidence (because people will notice if the evidence doesn’t warrant the conclusion); and it might help with other problems as well, like double-counting evidence due to many people stating the same conclusion.
I think there’s a weird thing where it’s best to *both* ramble more but *also* prune more? (something like, I think it’d be better if more people wrote off the cuff shortform thoughts, and also if more people dedicated some time to thinking through their positions more carefully.
I agree. OP is saying people should be more willing to make low-confidence predictions, whereas the type of rambling that people do too much of is information-sparse (taking too many words to say something simple). More rambling about meaningful but low-confidence claims, and less meaningless/redundant rambling.
Agreed; those are important considerations. In general, I think a risk for rationalists is to change one’s behaviour on complex and important matters based on individual arguments which, while they appear plausible, don’t give the full picture. Cf Chesterton’s fence, naive rationalism, etc.
This has to be carefully targeted to a select group of people or the world would be drowning in noise (even more than it already is). I think many people tend to be overconfident in their own ideas and already ramble more and filter less than they ideally should. Fear of being wrong might help counteract that somewhat so I’m wary of trying to remove that counterweight in someone before making sure to fix the overconfidence problem, unless you’re only targeting the intervention on people who already aren’t overconfident.
I think both overconfidence and underconfidence are widespread, so it’s hard to tell which advice would do more good. Maybe we can agree that people tend to over-share their conclusions and under-share their evidence? That seems plausible overall; advising people to shift toward sharing evidence might help address both underconfidence (because evidence feels safer to share) and overconfidence (because people will notice if the evidence doesn’t warrant the conclusion); and it might help with other problems as well, like double-counting evidence due to many people stating the same conclusion.
I think there’s a weird thing where it’s best to *both* ramble more but *also* prune more? (something like, I think it’d be better if more people wrote off the cuff shortform thoughts, and also if more people dedicated some time to thinking through their positions more carefully.
The Babble and Prune Sequence seems relevant here
I agree. OP is saying people should be more willing to make low-confidence predictions, whereas the type of rambling that people do too much of is information-sparse (taking too many words to say something simple). More rambling about meaningful but low-confidence claims, and less meaningless/redundant rambling.
Agreed; those are important considerations. In general, I think a risk for rationalists is to change one’s behaviour on complex and important matters based on individual arguments which, while they appear plausible, don’t give the full picture. Cf Chesterton’s fence, naive rationalism, etc.