The world of secondhand bookselling is abuzz with speculation about a peculiar trend: a flurry of bulk orders from mysterious buyers, some of which are suspected to be AI companies. This phenomenon has sparked concern and intrigue among booksellers, who are witnessing a surge in orders that defy typical patterns. The story begins with Stuart Manley, the co-owner of Barter Books in Northumberland, who noticed a strange combination of books in his orders. These orders, he noted, lacked the usual thematic grouping, appearing as a random assortment. Jim Shaughnessy, from MW Books in Ireland, echoed a similar experience, describing orders as sporadic and machine-driven, with no clear thematic connection. David Gower-Spence, owner of BookLovers of Bath, also reported a rush of orders from buyers, some of which have been linked to other sellers. The anonymity of these buyers adds to the mystery, with some orders seemingly coming from the same buyers under different aliases.
One of the most intriguing aspects of this story is the potential involvement of AI companies. The speculation arises from the fact that these orders often include books that are not typically sought after by AI training programs. For instance, the request for an Estonian translation of John le Carré’s The Mission Song or a specific edition of Warship magazine seems odd for AI training. The concern among booksellers is that these books, once acquired, might be scanned for their content before being pulped, raising ethical questions about the use of secondhand books for AI training. The Washington Post's revelation about Anthropic, an AI company, spending tens of millions of dollars on books for scanning, further fuels this speculation.
The AI industry's thirst for data is well-documented, and secondhand books offer a unique, unpolluted source of text. However, the practice of acquiring and destroying books for AI training raises ethical concerns. The online forum post expressing shame over the fate of these books highlights the emotional impact of the situation. The booksellers' struggle to understand the purpose of these erratic orders adds to the intrigue, especially when some of the requested titles are freely available online.
As the investigation continues, the secondhand bookselling community grapples with the implications of these bulk orders. The potential involvement of AI companies in this mystery not only disrupts the traditional trade but also raises important questions about the ethical use of data in the AI industry. The story serves as a reminder of the complex relationship between technology, data, and the preservation of cultural heritage.
