Proofs, Implications, and Models — LessWrong