check in
[Ultimately_Untrue_Thought.git] / notes / notes.txt
index d98f197..fe626a3 100644 (file)
@@ -2032,3 +2032,10 @@ https://reclaimthenet.org/threats-of-violence-censorship-for-womens-rights-group
 https://archive.vn/hQFh1
 
 pee fantasy: https://ericatgswapcaps.blogspot.com/2020/09/swap-clinic-favorite-thing.html
+
+https://www.reddit.com/r/TwoXChromosomes/comments/f0dfmt/has_anyone_ever_wrestled_with_guys_and_be/
+
+> Make a mistake, suffer a disaster, learn a lesson. Make two mistakes, suffer a disaster, learn half a lesson.
+https://twitter.com/spaceandgames/status/210229439395659776
+
+Oftentimes you genuinely happen to run out of time for a particular conversation, and that's fine: if your stopping algorithm is based on how busy you are that day and the expected information value of the remaining conversation, then we should expect to get accurate shared maps _eventually_, even if some topics get "unlucky" that you were busy that day. If your stopping algorithm is based on political convenience (you leave the conversation _because_ the other party made a good point that you're not allowed to admit in Berkeley), then we _don't_ converge to accurate shared maps!