Aumann's Agreement Theorem - History — LessWrong