Victoria Krakovna on AGI Ruin, The Sharp Left Turn and Paradigms of AI Alignment — LessWrong