“I underestimated safety research speedups from safe AI” by Dan Braun

LessWrong (30+ Karma) - “I underestimated safety research speedups from safe AI” by Dan Braun