AI4Math 中的淘金热：我们现在处于何处？
The Gold Rush in AI4Math: Where Are We Now?
浏览论文内容
中文总结 AI 辅助
该研究通过分析2026年3-8月的arXiv数学类论文，揭示AI辅助数学研究快速扩张但采用不均衡，组合数学、度量几何等领域及美中占比突出，OpenAI系统使用最频繁。
中文摘要 AI 辅助
人工智能（AI）的最新进展激发了人们对其在数学研究中应用的浓厚兴趣。一些人认为这是发现的重大机遇，而另一些人则对其对传统研究实践的影响提出了担忧。尽管存在广泛的争论，但关于 AI 在数学中实际使用情况的实证证据仍然有限。为了填补这一空白，我们收集了 2026 年 3 月 1 日至 8 月 20 日期间提交的所有 32944 篇 arXiv 论文，这些论文的主要或次要分类包含数学类。我们识别出 3575 篇明确披露作者使用 AI 的论文，其中 1712 篇涉及至少一项实质性数学贡献。我们的分析揭示了几个广泛的模式。首先，在研究期间，披露的 AI 使用量急剧增加，实质性使用量从 3 月占数学论文的 1.39%增长到 8 月 20 日的 14.09%。其次，实质性 AI 使用在各领域分布极不均衡：组合数学领域此类论文数量最多，而度量几何领域的实质性使用率最高。第三，实质性 AI 使用在地理上集中：按加权作者数量计算，美国和中国合计占公认国家权重的约三分之二。第四，AI 已被应用于未解决的研究问题：在与实质性使用相关的 717 个命名未解决问题记录中，71%根据作者的描述被标记为完全解决，其中猜想的证明比反例或反证更为常见。最后，AI 系统的使用也高度集中，OpenAI 系统出现频率最高，其次是 Anthropic。这些发现共同表明，AI 辅助数学研究正在迅速扩张，但仍处于采用的早期且不均衡的阶段。
英文摘要
Recent advances in artificial intelligence (AI) have sparked growing interest in its use for mathematical research. While some view this as a major opportunity for discovery, others have raised concerns about its impact on traditional research practices. Despite extensive debate, empirical evidence on how AI is actually being used in mathematics remains limited. To address this gap, we collected all 32,944 arXiv submissions posted between March 1 and August 20, 2026, whose primary or secondary categories included Mathematics. We identified 3,575 submissions that explicitly disclosed author use of AI, of which 1,712 involved at least one substantive mathematical contribution. Our analysis reveals several broad patterns. First, disclosed AI use increased sharply over the study period, with substantive use growing from 1.39% of Mathematics submissions in March to 14.09% through August 20. Second, substantive AI use is highly uneven across fields: Combinatorics has the largest number of such papers, while Metric Geometry has the highest substantive-use rate. Third, substantive AI use is geographically concentrated: under weighted author counts, the United States and China together account for about two-thirds of the recognized country weight. Fourth, AI is already being applied to open research problems: among 717 named open-problem records associated with substantive use, 71% are labeled as fully resolved based on the authors' descriptions, with proofs of the conjectured statement more common than counterexamples or disproofs. Finally, AI-system use is also highly concentrated, with OpenAI systems appearing most frequently, followed by Anthropic. Together, these findings suggest that AI-assisted mathematics is expanding rapidly but remains at an early and uneven stage of adoption.
