2026年5月，OpenAI 宣布其內部人工智慧模型成功為保羅·艾狄胥於1946年提出的「單位距離」問題找到了反例。這是歷史上首次由人工智慧提供具有重大數學意義的證明，該模型創新地運用了代數數論的技巧來解決這個難題。隨後，該模型又解決了更多艾狄胥提出的問題，這讓許多數學家認為人工智慧在數學能力上經歷了重大的相變，正深刻改變數學研究的進行方式。
這一連串突破的背後，英國數學家托馬斯·布魯姆功不可沒。他在2023年建立了一個專門收集艾狄胥遺留問題的網站，並隨後加入了評論功能，意外促成了一個高度民主化的線上社群。在這個平台上，無論是業餘愛好者、大學生，還是像陶哲軒這樣頂尖的數學家，都能夠在平等的基礎上交流合作，並開始利用公開的大型語言模型解決了一些較為簡單的艾狄胥問題。
隨著人工智慧能力的提升，數學界的生態正在發生根本性的轉變。對許多業餘愛好者而言，人工智慧不僅提升了他們解決問題的效率，也讓數學研究變得更加有趣且容易親近。然而，一些資深數學家則認為，既然機器已經能解決這些難題，繼續研究便失去了挑戰性與意義；這種趨勢甚至促使一些獲得菲爾茲獎等最高榮譽的頂尖數學家離開學術界，轉而投身於人工智慧科技公司。
In May 2026, OpenAI announced that its internal artificial intelligence model had successfully found a counterexample to the "unit distance" problem proposed by Paul Erdős in 1946. This marked the first time in history that an AI provided a proof of significant mathematical importance, innovatively applying techniques from algebraic number theory to tackle the challenge. The model subsequently solved even more problems posed by Erdős, leading many mathematicians to conclude that AI has undergone a major phase transition in its mathematical capabilities and is profoundly changing the way mathematical research is conducted.
Behind this series of breakthroughs, British mathematician Thomas Bloom played a crucial role. In 2023, he built a website dedicated to cataloging the unresolved problems left by Erdős and later added a commenting feature, unintentionally fostering a highly democratic online community. On this platform, amateurs, undergraduates, and top-tier mathematicians like Terence Tao were able to communicate and collaborate on an equal footing, eventually leading to the resolution of several simpler Erdős problems using publicly available large language models.
As artificial intelligence capabilities continue to improve, the ecosystem of the mathematical community is undergoing a fundamental transformation. For many hobbyists, AI not only increases their efficiency in problem-solving but also makes mathematical research more enjoyable and accessible. However, some seasoned mathematicians feel that since machines can now solve these puzzles, continuing to research them has lost its challenge and meaning; this trend has even prompted some top mathematicians, including a recipient of the highest honor like the Fields Medal, to leave academia and join AI technology companies instead.