• @[email protected]
    link
    fedilink
    79 days ago

    This actually helps AI, once his writing gets tagged as an attempt to confuse models it can be adapted to. Also because humans aren’t as good at random as they think they are it’s likely easy for an LLM to recognize.

    • @[email protected]
      link
      fedilink
      English
      49 days ago

      Exactly. Attention mechanisms excel at extracting signal from noise. This would simply reinforce that noise can come in this shape.