A lawsuit filed by several authors against Meta centers on Meta’s alleged use of pirated books for AI training data and the technical details of BitTorrent which was used to obtain them. Yesterday, Meta filed a motion for summary judgment, while countering the authors’ request to resolve the copyright claims in their favor. Meta’s request includes new information, including the revelation that its uploads of ‘pirate’ library data were roughly 30% of the data it downloaded.

  • pedroapero@lemmy.ml
    link
    fedilink
    English
    arrow-up
    15
    ·
    edit-2
    3 days ago

    Meta’s expert […] argued that […] if Meta shared small blocks of data, they would be unusable to the receiver.

    This is ridiculous. On a large torrent, a single piece can contain dozens of books. Pieces are contiguous unencrypted data. One piece contains several pages in any cases. What if I set my maximum ratio to 0.999, am I allowed to seed then?

    • RvTV95XBeo@sh.itjust.works
      link
      fedilink
      English
      arrow-up
      7
      ·
      3 days ago

      Even setting seed to infinite, if there’s just one other capable seeder, good odds no individual sends any other individual a full file.

      You’re just sending jibberish chunks everywhere, not your fault if someone assembles it all from multiple sources, right?