My response to AI 2027: https://t.co/hqqXLM5UZl The AI 2027 post is high quality, I encourage people to read it at https://t.co/84O5mfN5dh I argue a misaligned AI will not be able to win nearly as easily as the AI 2027 scenario assumes, because it greatly underrates our ability to protect ourselves, especially given the (pretty magical) technologies that the authors admit will be available in 2029 in their scenario.