Daily Tech Briefing
AI 科技速览

每天 5 分钟内学习 AI。获取最新的人工智能新闻,理解其重要性,并学习如何将其应用于您的工作。

AI 快讯
Hacker News · 2026/7/29 10:54:07
The Vanishing Page: AI Firms Scan Then Destroy Rare Book Editions

The Vanishing Page: AI Firms Scan Then Destroy Rare Book Editions

AI 中文解读
AI公司正在批量买书“喂”AI,却把原书直接销毁了!据法庭文件披露,Anthropic的“巴拿马计划”从2024年启动,用液压切割机切开书脊、高速扫描书页,再把纸质书粉碎回收。这些公司专挑2022年前出版的书,因为那时还没有AI生成内容,数据更“纯净”。中间商ISBNdb甚至能匿名处理百万册级别的订单,书店也不知道买家是谁。最让人意外的是,法院居然裁定这种做法算“合理使用”,理由是买了书就有权数字化,而且没有传播复制品。 这事的核心矛盾是:AI公司为了获取干净的训练数据,把人类珍贵的纸质文化遗产变成了一次性耗材。市面上已经有稀有书籍因此消失,据传有仅存个位数副本的绝版书也进了切割机。对我们普通人来说,这意味着二手书市场可能涨价,部分珍本彻底绝迹。同时,AI模型确实能因此学到更纯粹的文本知识——用“毁书”换来的智能提升,这笔账值不值,值得整个行业深思。
AI companies are purchasing physical books in bulk, including rare and out-of-print titles, then using high-speed machines to cut the spines and scan the pages before disposing of the originals, according to court documents and industry reports. The process targets pre-2022 books preferred as free of AI-generated content, with services facilitating anonymous bulk orders of up to one million volumes.A federal judge ruled the digitization of purchased books, accompanied by destruction of the physical copies, qualifies as fair use. Anthropic previously conducted a large-scale version of the practice under an internal effort known as Project Panama.The Practice and Legal BackgroundThe details resurfaced again after a July 26, 2026, post by the X account @HedgieMarkets described the activity, citing reports of rare books with few surviving copies entering the pipeline. 🦔AI companies are bulk-buying rare books, scanning them through high-speed machines that cut the spines off, and shredding the originals. A service called ISBNdb facilitates orders of up to a million books and keeps buyers anonymous. Pre-2022 books are premium because they're… pic.twitter.com/0YPd8N2oi5 — Hedgie (@HedgieMarkets) July 27, 2026The post referenced a bookseller’s account to 404 Media and statements from ISBNdb, a company that maintains a large book database and now brokers high-volume acquisitions for AI clients.In a June 2025 ruling in Bartz v. Anthropic, U.S. District Judge William Alsup of the Northern District of California held that digitizing legally purchased print books and using the resulting digital copies to train large language models was fair use. The judge noted that the process replaced the print copies with digital versions without adding new copies or redistributing them. Anthropic later settled related claims involving pirated digital books for $1.5 billion.Anthropic’s Project PanamaCourt filings unsealed in connection with a copyright lawsuit against Anthropic detailed Project Panama, which began in early 2024. An internal planning document stated that “Project Panama is our effort to destructively scan all the books in the world.” The same document added that “We use a ‘soft codename’ for it because we don’t want it to be known that we are working on this. This document is visible to all Anthropic employees, but you should avoid talking about it in public areas, and the fact that we are working on this should not be shared with anyone outside Anthropic.”Anthropic hired Tom Turvey, the former head of partnerships for Google Books, to obtain books. The company purchased volumes from used-book sellers including Better World Books and World of Books. A vendor proposal referenced converting 500,000 to two million books over a six-month period. The process used a hydraulic-powered cutting machine to remove pages, followed by industrial scanners, after which the physical materials were disposed of or recycled.Role of Brokers and Booksellers’ ReportsISBNdb markets physical books for AI training, stating on its site: “The world’s best AI training data is sitting on a shelf.” It emphasizes that print books from the pre-LLM era are free of AI-generated content. The company facilitates orders ranging from 1,000 to one million books and offers non-disclosure agreements. Its site states that “The optics problem is real. ‘AI company destroys two million books’ is not a headline that generates sympathy.”Booksellers in the Netherlands, Germany, Switzerland, and Spain have reported bulk orders for niche and specialized titles. Pieter de Vries of De Vries & De Vries in Haarlem received a request for nearly 3,000 titles from a Singapore-based firm. A German seller noted overnight orders from the Canadian company Zoom Books for unrelated academic works. Dealers stated the purchases did not match typical collector or resale patterns.Destructive scanning is used because it is faster and less expensive than non-destructive methods. The physical originals ar
分享
阅读原文