Complexity and Intelligence — LessWrong