An angle of attack on Open Problem #1 — LessWrong