Using Intellectual Processes to Combat Bias — LessWrong