cross-posted from: https://lemmy.world/post/1330512
Below are direct quotes from the filings.
OpenAI
As noted in Paragraph 32, supra, the OpenAI Books2 dataset can be estimated to contain about 294,000 titles. The only “internet-based books corpora” that have ever offered that much material are notorious “shadow library” websites like Library Genesis (aka LibGen), Z-Library (aka B-4ok), Sci-Hub, and Bibliotik. The books aggregated by these websites have also been available in bulk via torrent systems. These flagrantly illegal shadow libraries have long been of interest to the AI-training community: for instance, an AI training dataset published in December 2020 by EleutherAI called “Books3” includes a recreation of the Bibliotik collection and contains nearly 200,000 books. On information and belief, the OpenAI Books2 dataset includes books copied from these “shadow libraries,” because those are the most sources of trainable books most similar in nature and size to OpenAI’s description of Books2.
Meta
Bibliotik is one of a number of notorious “shadow library” websites that also includes Library Genesis (aka LibGen), Z-Library (aka B-ok), and Sci-Hub. The books and other materials aggregated by these websites have also been available in bulk via torrent systems. These shadow libraries have long been of interest to the AI-training community because of the large quantity of copyrighted material they host. For that reason, these shadow libraries are also flagrantly illegal.
This article from Ars Tecnica covers a few more details. Filings are viewable at the law firm’s site here.
Mentioning this since the project Anna’s Archive compiles several datasets and their corresponding torrents.
Anna’s Archive, whose aim is to “archive all the books in the world, and make them widely accessible,” pulls from a number of shadow library sources; the project provides its own torrent links (via Tor) for Library Genesis, Z-lib, Internet Archive, among others, plus Library Genesis’s torrents. In the datasets linked below, you can click on a given source and find its onion site or the torrents provided by the shadow library itself (in the case of Library Genesis, for example).
Anna’s Archive datasets
Sources include
Thanks a lot.