The Flaw in the Paperclip Maximizer Thought Experiment — LessWrong