x
Tall tales and long odds — LessWrong