[LINK] Cantor's theorem, the prisoner's dilemma, and the halting problem — LessWrong