setVeryLoud(true);

  • 0 Posts
  • 5 Comments
Joined 26 days ago
cake
Cake day: May 10th, 2025

help-circle
  • setVeryLoud(true);@lemmy.worldtoFuck AI@lemmy.worldOn AI
    link
    fedilink
    arrow-up
    1
    ·
    edit-2
    11 hours ago

    We should also be specific in what we oppose.

    “Artificial Intelligence” is probably one of the broadest terms in common use today.

    IMO, this is what we should oppose:

    • Diffusion, tensor and GAN-based Image and video generation, no exceptions, they serve zero practical purpose
    • Voice duplication, GAN-based voice generation is ok and is a good thing for accessibility
    • Non-voice GAN-based audio generation
    • Black box LLMs, LLMs can be genuinely useful tools but they HAVE to be trained ethically in a provable way that matches the OSI’s definition of FLOSS

    I might have missed some, but IMO, large language models and voice generation can be done ethically and are ok in very narrow contexts. Both of these are tools and should never be used as a final product.