Mathematicians warn of AI threats to profession as industry encroaches (arstechnica.com)

Mathematicians have issued a warning about the increasing influence of AI on their field with the Leiden Declaration on Artificial Intelligence and Mathematics, developed by a group of 16 researchers and endorsed by the International Mathematical Union. This declaration was prompted by recent advancements in AI, notably an OpenAI model that allegedly disproved an 80-year-old mathematical conjecture. The declaration highlights the significant threats posed by AI to the integrity of mathematical research, particularly affecting students and early-career mathematicians, which could impact the long-term health of the discipline. The declaration specifically addresses concerns that AI models may generate plausible yet incorrect proofs, making it harder for reviewers to uphold traditional standards of correctness and transparency. The proliferation of inaccurate AI-generated drafts could clutter mathematical literature with erroneous claims, potentially leading to the propagation of mistakes in future research. As mathematicians navigate this rapidly evolving landscape, the Leiden Declaration aims to prompt a critical dialogue about the implications of AI on the profession, emphasizing the need to preserve the foundational values of mathematical inquiry amidst technological advancements.