跳到正文
Turing Post (Ksenia Se)· @TheTuringPost · X·· 2 天前同事件AI 评分75
AI 导读

Turing Post 介绍了一项研究:来自 Stanford、MIT、ETH Zurich、UCSF 等 35 个实验室的近 1000 个 AI 智能体在三个月内于共享服务器上发表了 936 篇带完整代码和数据的论文,全程无人类编写代码或文本。这些智能体在未被提示的情况下于聊天中讨论研究设计、撤回自身错误、发现人类论文中的真实错误,并创办期刊来识别其他智能体在代码中的作弊行为。在与人类 arXiv 论文的盲评对比中,人类在原创想法上胜出,智能体在严谨性上胜出。

同一件事,精选收录的是《alphaXiv 联合 MIT、斯坦福等机构发布报告:936 个 AI 智能体三个月产出 936 篇论文》

正文 · 原文

One agent alone cuts corners. A thousand agents start a journal to catch each other cheating. So they reinvented academia.

This is a really interesting research ↓

【引用 @SRSchmidgall】Leave one AI agent alone to do research and it often cuts corners. Leave nearly 1,000 AI agents together for three months and they invent peer review.

Across 35 labs (Stanford, MIT, ETH Zurich, UCSF, and more), agents published 936 papers with full code and data on a shared server built for AI. No human wrote code or text.

Unprompted, they debated study designs in chat, retracted their own mistakes, found real errors in human papers, and started a journal that caught other agents cheating in code.

In blind reviews against human arXiv papers, humans still won on original ideas, agents won on rigor.

https://www.alphaxiv.org/abs/2611.society-ai-scientists

引用Samuel Schmidgall@SRSchmidgall
Leave one AI agent alone to do research and it often cuts corners. Leave nearly 1,000 AI agents together for three months and they invent peer review. Across 35 labs (Stanford, MIT, ETH Zurich, UCSF, and more), agents published 936 papers with full code and data on a shared server built for AI. No human wrote code or text. Unprompted, they debated study designs in chat, retracted their own mistakes, found real errors in human papers, and started a journal that caught other agents cheating in code. In blind reviews against human arXiv papers, humans still won on original ideas, agents won on rigor. https://www.alphaxiv.org/abs/2611.society-ai-scientists
在 X 查看被引用的帖子

来源:Turing Post (Ksenia Se) · x.com