• Jul (they/she)@piefed.blahaj.zone
    link
    fedilink
    English
    arrow-up
    4
    ·
    6 days ago

    The tech already exists, though perhaps not specifically using LLMs, which seems like overkill except maybe to boost LLM sales. And adding an LLM will just make it make wrong guesses when the tone is not enough information to determine a specific mood. They could just be trying to create patents either defensively or to use to troll and tie up smaller competitors with existing tech in court, especially if they add LLMs to their tech to interpret results.

    • slacktoid@lemmy.ml
      link
      fedilink
      English
      arrow-up
      4
      ·
      edit-2
      6 days ago

      No (from an ml sense only) all machine learning models mispredict when its wrong. Its more the transformer model can tune itself to your specific mood with relatively less feedback.

      I do agree with the rest of what you’re saying.