AI Safety Newsletter #1 [CAIS Linkpost] — LessWrong