What is LessWrong explicitly trying to optimize?(And what percentage of its structure currently serves that goal?) — LessWrong