Attempts to work around Goedel's theorem by using randomness — LessWrong