Mathematician Xiaoyu He published an exposition arguing that rapid progress in AI, including recent breakthroughs in automated theorem proving, warrants the mathematics community taking AI existential-risk arguments seriously. The essay lays out concerns about AI takeover scenarios, recursive self-improvement, and alignment difficulties, alongside the cognitive biases the author says may prevent adequate risk mitigation. It frames the debate as relevant not only to AI safety researchers but to mathematicians assessing the future of their own profession.
