David Scott Krueger (formerly: capybaralet) comments on Soon: a weekly AI Safety prerequisites module on LessWrong