Thoughts on Iason Gabriel’s Artificial Intelligence, Values, and Alignment — LessWrong