部分数学家呼吁抵制 OpenAI,AI 生成的数学证明涌入数学领域

来源:The Decoder:AI News(RSS) 2026年10月8日 02:17 AIHOT 评分:81 精选
摘要:AHM 组织在 Terence Tao 主持下发表声明,呼吁抵制 OpenAI,称其一次性发布 700 多个 AI 生成证明文件是展示力量而非学术行为。此前 OpenAI 宣称内部模型一个月内解决逾 100 个开放数学问题,约 8000 个测试问题成功率约 5%,平均每个耗时 3 小时 GPT-Pro 级算力。
推荐理由:材料汇集了数学界对 OpenAI 批量发布 AI 证明的抵制声明与 Tao、Aaronson 等多方反应,便于了解争议全貌与社区分歧。
# 部分数学家呼吁抵制 OpenAI,AI 生成的数学证明涌入数学领域 > 原标题:Some mathematicians call for OpenAI boycott after AI-generated proofs flood their field - 来源:The Decoder:AI News - 发布时间:2026-10-08T18:17:07.000Z - AIHOT:https://aihot.news/items/wf099r75jv1tzyhxa2j2o8n1p - 原文:https://the-decoder.com/some-mathematicians-call-for-openai-boycott-after-ai-generated-proofs-flood-their-field/ ## 摘要 AHM 组织在 Terence Tao 主持下发表声明,呼吁抵制 OpenAI,称其一次性发布 700 多个 AI 生成证明文件是展示力量而非学术行为。此前 OpenAI 宣称内部模型一个月内解决逾 100 个开放数学问题,约 8000 个测试问题成功率约 5%,平均每个耗时 3 小时 GPT-Pro 级算力。 ## 推荐理由 材料汇集了数学界对 OpenAI 批量发布 AI 证明的抵制声明与 Tao、Aaronson 等多方反应,便于了解争议全貌与社区分歧。 ## 正文 · 中文译文 担任该小组主席的菲尔兹奖得主陶哲轩(Terence Tao)[以客座文章的形式在其博客上分享了这份声明](https://terrytao.wordpress.com/2026/10/07/ahm-statement-on-openais-october-6-release-of-mathematical-documents/),同时还在 Mastodon 上勾勒出一个“数学 2.0”时代,在这个时代,解决问题不应再是该学科的主要焦点。复杂性理论家斯科特·阿伦森(Scott Aaronson)在[其博客](https://scottaaronson.blog/?p=10169)上将这一局面描述为“数学末日”(Mathocalypse)。 AHM 的声明使一场[持续数月的争论](https://the-decoder.com/ai-keeps-cracking-unsolved-math-problems-and-mathematicians-have-mixed-feelings/)进一步升级。OpenAI 此前已引发轩然大波,声称其内部模型在一个月内解决了[100 多个未解数学问题](https://the-decoder.com/openai-says-its-internal-model-solved-over-100-long-standing-math-problems-after-just-a-month-of-training/?cmpscreencustom=1),其中包括针对[纳维-斯托克斯千禧年难题](https://the-decoder.com/openai-reportedly-closes-in-on-solving-the-hodge-conjecture-its-second-millennium-prize-problem/)的一种解法。面对日益增多的批评,[OpenAI 在高等研究院设立了顾问小组 AGMAI](https://the-decoder.com/openai-says-its-internal-model-solved-over-100-long-standing-math-problems-after-just-a-month-of-training/),不过该小组明确对公司内部研究的节奏没有发言权。 ## AHM 将此次发布视为一种武力展示 AHM 表示,OpenAI 已经无视了 AGMAI 建议的核心前提,即不应在内部模型上测试高等数学问题。声明中写道:“数学家们并未要求做这项工作”,并将一次性发布 700 多个文件描述为“不是学术的展示,而是权力的展示”。 声明的开篇段落还提到了 OpenAI 正在全球范围内应对的众多版权诉讼,虽未明说,但暗示该公司之所以能取得这些成果,是因为它训练时使用了数学家的作品,无论合法与否。 同样的争议在 OpenAI 最近宣布解决[纳维-斯托克斯问题](https://the-decoder.com/openai-researcher-allegedly-pressured-mathematician-to-drop-anthropic-co-author-from-math-breakthrough-paper/)时再次浮出水面,而就在不久之前,两位数学家正准备展示他们自己借助 AI 得出的解法。这两人在工作中使用了 ChatGPT,OpenAI 否认了有关其利用[这些交互中的数据](https://the-decoder.com/ai-labs-have-a-data-trust-problem-that-their-policies-havent-solved/)来训练自家系统并更快得出解法的猜测。 阿伦森将此与他所称的“Anthropic 模式”进行了对比。OpenAI 一次性发布原始 AI 证明草稿,引发了一场争相研读的竞赛。[Anthropic 则与两位算法研究人员合作](https://www.anthropic.com/research/formalizing-fermats-last-theorem),其 AI 模型为推翻两个已有二十年历史的猜想提供了关键思路。研究人员获得了报酬,并撰写了一份人类可以看懂的证明版本。 两种方法都有缺点。OpenAI 的做法让社区免费承担起使证明可读的吃力不讨好的工作。Anthropic 的做法则让一家私营公司决定哪些数学家可以充当某项成果的“代言人”。 ## AI 破解数学问题的速度超过了任何人验证结果的速度 陶哲轩此前曾与[其他 24 位菲尔兹奖得主](https://the-decoder.com/leading-mathematicians-fear-ai-is-making-their-field-dumber-and-warn-the-rest-of-us-is-next/?cmpscreencustom=1)——包括彼得·舒尔茨(Peter Scholze)、玛丽娜·维亚佐夫斯卡(Maryna Viazovska)和马丁·海雷尔(Martin Hairer)——共同签署了一份声明,警告 AI 行业的目标与数学的目标之间存在“严重错位”。他们认为,解决问题只是实现该学科真正目标——概念性理解——的工具,而以越来越快的速度批量产出已解决的问题,可能“摧毁肥沃的土壤,而不是为新思想注入活力”。 签署者还批评了匆忙发布 AI 生成的解法、却缺乏适当撰写说明或引用相关先前工作的做法,称这引发了“严重的署名和剽窃问题”。 在 [Mastodon](https://mathstodon.xyz/@tao/117395267721642920) 上,陶哲轩如今更进一步。在传统数学中,对长期未解问题的证明会带来讲座、研讨会、新的合作,并被纳入教科书,吸引年轻研究者去理解并为其提供背景。 如今他看到的恰恰相反。问题正被 AI 用户自主解决,而这些人对该领域毫无兴趣,也无法充分理解结果以回答问题或做报告,由此产生的研讨会和合作远少于传统突破本应带来的数量。研究者也在隐瞒有前景的方向,担心竞争者抢先给出解答。 对陶哲轩而言,这种损害不可逆转。一个问题一旦被认为已解决,就无法再变回未解决。正如他所说,即便只是知道存在某个解答,也可能“污染”对其他可能带来更多洞见的路径的探索。解答正被大规模“收割”,使整个数学分支的肥沃程度不如从前。 陶哲轩主张,“数学 2.0”时代必须停止把解题当作衡量进步的主要标准,转而重视解释、社群建设以及开辟新的研究方向,并相应改变对训练、发表和职业晋升的评判方式。 这建立在陶哲轩 2024 年提出的“[工业数学](https://the-decoder.com/renowned-mathematician-terence-tao-envisions-ai-powered-industrial-scale-mathematics/)”愿景之上,不过在那一版本中,人类仍掌握节奏。AI 本应像引擎辅助棋手那样辅助数学家,使广泛而相对浅层的研究成为可能,并补充人类专家的深层工作。如今,陶哲轩认为这些角色已经反转。AI 正在解决深层问题,而人类事后梳理结果。[与许多其他专家一样](https://the-decoder.com/top-ai-experts-badly-underestimated-how-fast-the-field-is-moving-study-finds/),陶哲轩显然[严重低估了 AI 的发展速度](https://the-decoder.com/math-genius-terence-tao-says-that-ai-still-cant-smell-bad-math/?cmpscreencustom=1)。 [![](https://the-decoder.com/wp-content/uploads/2026/10/gowers_x-1.png)](https://the-decoder.com/wp-content/uploads/2026/10/gowers_x-1.png) 研究者之间流传着一些黑色幽默。[蒂莫西·高尔斯](https://the-decoder.com/top-mathematicians-say-llms-are-strong-calculators-but-poor-creative-thinkers/)本人也是 AGMAI 成员,他分享了 Nicolo Cesa-Bianchi 在 X 上的一篇帖子,后者在米兰大学研究机器学习算法。 | 图片:X 截图 ## 三小时的 AI 算力如今可媲美一位数学家的整个职业生涯 Aaronson 通过妻子 Dana Moshkovitz 的经历,展示了这对个体研究者意味着什么。她整个职业生涯都投入到理论计算机科学中的一个未解问题——唯一游戏猜想。而这次发布包含了一个对该猜想的证明声明。 在发布当晚发给 Aaronson 的短信中,她写道:“感觉像是某个嗑了药的人写的。”文件的大部分内容都不清晰。论文引用了大量工作,却没有解释它们为何适用,尽管早先的结果本应排除这种可能。 “基本上,这篇论文写得糟糕透顶,没有 AI 帮助根本读不下去,”Moshkovitz 说。该证明依赖一个全新的、怪异的构造,她将其描述为“某种外星式的疯狂”。 Aaronson 还列举了复杂性理论、数论和算法研究中的其他结果,包括对若干尚未解决的克雷千年问题取得的局部进展。在他看来,其中任何一项单独拿出来都足以跻身年度顶级成果之列。 不过,此次发布中有一个领域明显缺席。密码学。Aaronson 援引匿名消息源报道称,AI 公司如今正在悄悄探究[密码协议与基础原语中的弱点](https://the-decoder.com/ai-math-breakthroughs-have-ethereum-researchers-debating-how-fast-wallet-security-could-collapse/?cmpscreencustom=1)。总体上共测试了约 8,000 个问题,成功率约为百分之五,每个问题在 GPT-Pro 级别下平均消耗三小时算力。所使用的模型很可能也是 OpenAI 当前的内部模型,该模型可能在未来几个月内向付费 ChatGPT 客户开放。 ## 数学家们对抵制是否有帮助意见不一 Tao 博客文章下的评论显示该领域分歧之大。几位评论者支持抵制,并提议研究人员停止使用 OpenAI 产品,同时呼吁更好地保护 arXiv 等预印本服务器,防止训练数据被大规模收集。 另一些人认为 AHM 的立场不切实际,认为 OpenAI 不会停止数学方面的工作,而且公开成果总比私下成果好。还有一位评论者问道,当作者和其他人都没有完全理解证明时,一个问题是否真的能算作已解决。就 Navier-Stokes 而言,目前尚不清楚这一结果是否真的加深了任何人对流体力学的理解。 [AGMAI 本身](https://agmai.org/#openai-release)采取的立场比 AHM 更为圆融,将该发布视为第一步,对结果的数学理解才刚刚开始。尽管如此,该组织警告称,数学研究的未来不能是逐一消化 AI 实验室产出的结果。 该组织写道:“数学家必须能够提出自己的问题、发展自己的方法,并探索那些未被选为 AI 系统能力示例的方向。公平地获取强大的研究工具和充足的计算资源,对于这种自由至关重要。” ## 正文 · 原文 Fields Medalist Terence Tao, who chairs the group, [shared the statement as a guest post](https://terrytao.wordpress.com/2026/10/07/ahm-statement-on-openais-october-6-release-of-mathematical-documents/) on his blog, while also outlining a "Math 2.0" era on Mastodon in which solving problems should no longer be the main focus of the discipline. Complexity theorist Scott Aaronson described the situation as a "Mathocalypse" on [his blog](https://scottaaronson.blog/?p=10169). The AHM statement escalates a [months-long debate](https://the-decoder.com/ai-keeps-cracking-unsolved-math-problems-and-mathematicians-have-mixed-feelings/). OpenAI had already caused a stir by claiming an internal model solved [more than 100 open math problems](https://the-decoder.com/openai-says-its-internal-model-solved-over-100-long-standing-math-problems-after-just-a-month-of-training/?cmpscreencustom=1) in a single month, including an approach to the [Navier-Stokes Millennium Problem](https://the-decoder.com/openai-reportedly-closes-in-on-solving-the-hodge-conjecture-its-second-millennium-prize-problem/). In response to growing criticism, [OpenAI set up the advisory group AGMAI](https://the-decoder.com/openai-says-its-internal-model-solved-over-100-long-standing-math-problems-after-just-a-month-of-training/) at the Institute for Advanced Study, though the group explicitly has no say over the pace of the company's internal research. ## AHM sees the release as a show of force The AHM says OpenAI has already ignored the central premise of AGMAI's advice, which held that advanced math problems shouldn't be tested on internal models. "Mathematicians did not ask for this work to be done," the statement reads, describing the release of more than 700 files at once as "not a demonstration of scholarship, but a demonstration of power." The statement's opening paragraph also brings up the many copyright lawsuits OpenAI is fighting around the world, suggesting without saying so directly that the company could only achieve these results because it trained on mathematicians' work, whether legally or not. That same dispute surfaced when OpenAI recently announced a solution to the [Navier-Stokes problem](https://the-decoder.com/openai-researcher-allegedly-pressured-mathematician-to-drop-anthropic-co-author-from-math-breakthrough-paper/) shortly before two mathematicians could present their own AI-assisted solution. The two had used ChatGPT in their work, and OpenAI denied suspicions that it had used [data from those interactions](https://the-decoder.com/ai-labs-have-a-data-trust-problem-that-their-policies-havent-solved/) to train its own system and reach a solution faster. Aaronson contrasts this with what he calls the "Anthropic model." OpenAI releases raw AI proof drafts in one batch, triggering a race to work through them. [Anthropic instead partnered with two algorithm researchers](https://www.anthropic.com/research/formalizing-fermats-last-theorem) whose AI model supplied the key idea for disproving two decades-old conjectures. The researchers received compensation and wrote a version of the proof that humans could follow. Both approaches have drawbacks. OpenAI's method leaves the community doing the thankless work of making proofs readable, for free. Anthropic's method lets a private company choose which mathematicians get to serve as "emissaries" for a result. ## AI is cracking open math problems faster than anyone can verify the results Tao previously joined [24 other Fields Medalists](https://the-decoder.com/leading-mathematicians-fear-ai-is-making-their-field-dumber-and-warn-the-rest-of-us-is-next/?cmpscreencustom=1), including Peter Scholze, Maryna Viazovska, and Martin Hairer, in signing a statement warning of a "severe misalignment" between the AI industry's goals and those of mathematics. Problem-solving, they argue, is only a tool for the discipline's real goal of conceptual understanding, and mass-producing solved problems at an ever-faster pace could "destroy fertile ground instead of breathing life into new ideas." The signatories also criticize rushed announcements of AI-generated solutions without proper write-ups or citations of relevant prior work, saying this raises "severe attribution and plagiarism questions." On [Mastodon](https://mathstodon.xyz/@tao/117395267721642920), Tao now goes further. In traditional mathematics, proofs of long-standing open problems would lead to talks, workshops, new collaborations, and inclusion in textbooks, attracting young researchers who would work to understand and contextualize them. Now he sees the opposite. Problems are being solved autonomously by AI users who have no interest in the field and can't understand the results well enough to answer questions or give talks, producing far fewer seminars and collaborations than traditional breakthroughs would. Researchers are also holding back promising directions for fear that competitors will scoop the solution. For Tao, the damage is irreversible. Once a problem is considered solved, it can't be made unsolved again. Even knowing a solution exists can "contaminate," as he puts it, the search for other approaches that might yield further insights. Solutions are being "harvested" on a large scale, leaving entire branches of mathematics less fertile than before. Tao argues that the "Math 2.0" era must stop treating problem-solving as its main measure of progress and start valuing explanation, community-building, and the opening of new research directions, with corresponding changes to how training, publication, and career advancement are judged. This builds on Tao's earlier vision of "[industrial mathematics](https://the-decoder.com/renowned-mathematician-terence-tao-envisions-ai-powered-industrial-scale-mathematics/)" from 2024, though humans still set the pace in that version. AI was supposed to assist mathematicians much as engines assist chess players, enabling broad, relatively superficial research that complemented the deeper work of human experts. Today, Tao sees those roles reversed. AI is solving the deep problems while humans work through the results afterward. [Like many other experts](https://the-decoder.com/top-ai-experts-badly-underestimated-how-fast-the-field-is-moving-study-finds/), Tao apparently [badly underestimated the pace of AI development](https://the-decoder.com/math-genius-terence-tao-says-that-ai-still-cant-smell-bad-math/?cmpscreencustom=1). [![](https://the-decoder.com/wp-content/uploads/2026/10/gowers_x-1.png)](https://the-decoder.com/wp-content/uploads/2026/10/gowers_x-1.png) Researchers share some gallows humor. [Timothy Gowers](https://the-decoder.com/top-mathematicians-say-llms-are-strong-calculators-but-poor-creative-thinkers/), himself an AGMAI member, shares a post on X by Nicolo Cesa-Bianchi, who studies machine learning algorithms at the University of Milan. | Image: Screenshot X ## Three hours of AI compute can now rival a mathematician's entire career Aaronson shows what this means for individual researchers through the experience of his wife, Dana Moshkovitz, who has devoted her entire career to the Unique Games Conjecture, an open problem in theoretical computer science. The release includes a claimed proof of it. In text messages to Aaronson on the night of the release, she wrote, "It feels like something written by someone who's on psychedelics." Much of the document was unclear. The paper cited numerous works without explaining why they applied, even though earlier results should have ruled that out. "Basically the paper is so horribly written that it's impossible to read it without AI help," Moshkovitz said. The proof relied on an entirely new, bizarre construction that she described as "some alien craziness." Aaronson lists other results from complexity theory, number theory, and algorithm research, including partial progress on several remaining Clay Millennium Problems. Any one of them, in his view, would have ranked among the year's top results on its own. One area is notably absent from the release, however. Cryptography. Citing unnamed sources, Aaronson reports that AI companies are now quietly probing [weaknesses in cryptographic protocols](https://the-decoder.com/ai-math-breakthroughs-have-ethereum-researchers-debating-how-fast-wallet-security-could-collapse/?cmpscreencustom=1) and primitives. About 8,000 problems were tested overall, with a success rate of roughly five percent and an average of three hours of compute at GPT-Pro level per problem. The model used was likely OpenAI's current internal one as well, which could ship to paying ChatGPT customers in the coming months. ## Mathematicians disagree over whether a boycott will help Comments on Tao's blog post show how divided the field is. Several commenters back the boycott and propose that researchers stop using OpenAI products while also calling for better protection of preprint servers like arXiv against mass collection of training data. Others consider the AHM's position unrealistic, arguing that OpenAI won't stop working on mathematics and that public results are better than private ones. Another commenter asks whether a problem can really count as solved when neither the authors nor anyone else fully understands the proof. In the case of Navier-Stokes, it remains unclear whether the result actually deepens anyone's understanding of fluid dynamics. [AGMAI itself](https://agmai.org/#openai-release) takes a more diplomatic stance than the AHM, viewing the release as a first step with the mathematical understanding of the results only now beginning. Still, the group warns that the future of math research can't consist of working through results produced by AI labs. "Mathematicians must be able to formulate their own questions, develop their own approaches, and explore directions that have not been selected as examples of an AI system's capabilities. Equitable access to powerful research tools and adequate computational resources are essential to that freedom," the organization writes.