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 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.