Chapter 16: Lateral Thinking — LessWrong