The Legacy of Computer Science — LessWrong