Comment on Judge orders Anna’s Archive to delete scraped data; no one thinks it will comply

aeronmelon@lemmy.world ⁨2⁩ ⁨days⁩ ago

• db2@lemmy.world ⁨2⁩ ⁨days⁩ ago

  Couldn’t they make an argument with that pointing out that they’re being unjustly targeted because they’re smaller and easier to pick on?

  source

  • SolacefromSilence@fedia.io ⁨2⁩ ⁨days⁩ ago

    No one cares if they're small or unjustly picked on. If they want to beat the charges, they need to announce their own AI trained on the data.

    source

    • tempest@lemmy.ca ⁨2⁩ ⁨days⁩ ago

      It would make me laugh if they could train an LLM that could only regurgitate content verbatim

      source

      • ilinamorato@lemmy.world ⁨2⁩ ⁨days⁩ ago

        Well, it’s not an LLM, but “AI” doesn’t have a defined meaning, so from that perspective they kind of already did.

        source
      • Dran_Arcana@lemmy.world ⁨1⁩ ⁨day⁩ ago

        en.wikipedia.org/wiki/Markov_chain

        Before the advent of AI, I wrote a slack bot called slackbutt that made Markov chains of random lengths between 2 and 4 out of the chat history of the channel. It was surprisingly coherent. Making an “llm” like that would be trivial.

        source

        • -> View More Comments
      • Natanael@infosec.pub ⁨1⁩ ⁨day⁩ ago

        It’s actually kinda easy. Neural networks are just weirder than usual logic gate circuits. You can program them just the same and insert explicit controlled logic and deterministic behavior.

        The reason that doesn’t work well with a traditional LLM despite the node properties being well known is because of how intricately interwoven and mutually dependent all the different parts of the network is. You can’t just arbitrarily edit anything or insert more nodes or replace logic, you don’t know what you might break. It’s easier to place inserted logic outside of the LLM network and train the model to interact with it (“tool use”).

        source
    • panda_abyss@lemmy.ca ⁨2⁩ ⁨days⁩ ago

      We’ve created a state of the art model in recall and training data idempotency!

      source
    • UnderpantsWeevil@lemmy.world ⁨2⁩ ⁨days⁩ ago

      If they want to beat the charges, they need to announce their own AI trained on the data several billion in Series A investment funding.

      source
    • db2@lemmy.world ⁨2⁩ ⁨days⁩ ago

      I hope they call it FUAI.

      source
• evol@lemmy.today ⁨2⁩ ⁨days⁩ ago

  rebrand into Anna AI !

  source