The short-term personal predictions you mention are very useful (at least for the person making them), because you can use them to calibrate yourself faster. (

The short-term personal predictions you mention are very useful (at least for the person making them), because you can use them to calibrate yourself faster. (If you want to get better at predicting, waiting 10 years for a 2021 prediction doesn't help as much.) It can also be fun to put estimates on other people's personal predictions based on their historical underconfidence/overconfidence. It has been suggested that these personal predictions should be in a separate category, and that might happen. Also, these articles on PB might be of interest: http://lesswrong.com/lw/7z9/1001_predictionbook_nights/ http://lesswrong.com/lw/7z9/1001_predictionbook_nights/ http://lesswrong.com/r/discussion/lw/8dx/predictionbook_a_short_note/ http://lesswrong.com/r/discussion/lw/8dx/predictionbook_a_sh...