“Reflections on my first year of AI safety research” by Jay Bailey | LessWrong (30+ Karma) | Podwise