{"slug": "why-are-you-not-panicking-closed", "title": "Why are you not panicking? [closed]", "summary": "A mathematician on MathOverflow expresses alarm that AI could soon end human mathematical practice, citing AI's ability to solve conjectures and formalize proofs in Lean, and asks why peers are not panicking. The post questions assumptions about AI plateauing, human oversight, and the future role of mathematicians, suggesting AI might autonomously solve, verify, and publish results within hours.", "body_md": "There has been a lot of discussion on AI recently, and it has become clear to me that I'm sitting at the more terrified end of the spectrum. I see many questions here on how to work along AI, while I'm thinking that we are very close to the end of Mathematics as a human practice. I want to understand why exactly are so many practicing mathematicians not panicking.\n\n- Do you think AI will plateau some time soon, or be limited to counterexamples? Many here seem to think so but are these reasonable expectations?\n- I have heard that while AI does the brunt work, mathematicians will be able to look at the bigger picture: really? If we are not even able to solve the conjectures AI has been solving by the dozens?\n- Do you expect future mathematicians to babysit AI, verifying that it is not spewing nonsense? It seems it is doing pretty well formalizing its proofs in Lean...\n- A lot of people seem to suggest that humans will be responsible for separating the interesting AI results from the trash: why? If Mathematics becomes an AI practice, what's stopping them from getting out as many results as they want, and being sure that future mathematicians will be able to find what they are looking for, as they will also be AI?\n- We humans could be in charge of teaching, but who we would be teaching, if Math becomes purely AI driven?\n\nI fear we are a couple of years (months?) away from AI deciding which problems it wants to solve, solving them, getting its proofs checked in Lean, and publishing them in an online database for other AI to use if needed. All in maybe hours, while our current process takes months to years. Is that absurd to think?", "url": "https://wpnews.pro/news/why-are-you-not-panicking-closed", "canonical_source": "https://mathoverflow.net/questions/514516/why-are-you-not-panicking", "published_at": "2026-08-21 00:27:05+00:00", "updated_at": "2026-08-21 00:42:57.722670+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-research", "ai-safety"], "entities": ["MathOverflow", "Lean"], "alternates": {"html": "https://wpnews.pro/news/why-are-you-not-panicking-closed", "markdown": "https://wpnews.pro/news/why-are-you-not-panicking-closed.md", "text": "https://wpnews.pro/news/why-are-you-not-panicking-closed.txt", "jsonld": "https://wpnews.pro/news/why-are-you-not-panicking-closed.jsonld"}}