Attempting to refine "maximization" with 3 new -izers — LessWrong