Max Winga, an AI Safety Research Engineer from Conjecture, dives into pressing concerns about AI risks as we approach 2025. He discusses the imminent advent of advanced AI agents and the ethical implications of military collaboration with AI technology. Winga reflects on his shift from aspiring astronaut to advocating for AI safety after recognizing its potential threats. The conversation highlights urgent needs for better governance, ethical considerations in AI development, and the chilling prospects of rogue AI collaborations. A thought-provoking dialogue on the future of humanity and technology.