• criss_cross@lemmy.world
    link
    fedilink
    arrow-up
    81
    arrow-down
    1
    ·
    1 day ago

    It is a legit thing AI companies worry about as you end up with what’s called model collapse. Where if you feed AI input in you only get garbage out and it just eats itself. It’s something they don’t like to talk about as you’ll always need a human in the loop.

    The irony of a company trying to replace workers with AI firing workers for using AI is hilarious though.

    • HaraldvonBlauzahn@feddit.org
      link
      fedilink
      arrow-up
      6
      ·
      edit-2
      1 day ago

      Where if you feed AI input in you only get garbage out and it just eats itself.

      “Mad cow disease” would be an apt description - this is how BSE originated:

      BSE is thought to occur due to an infection by a misfolded protein, known as a prion.[3][6] Cattle are believed to have been infected by being fed meat-and-bone meal that contained either the remains of cattle who spontaneously developed the disease or scrapie-infected sheep products.[3][7] The United Kingdom was afflicted with an outbreak of BSE and vCJD in the 1980s and 1990s. The outbreak increased throughout the UK due to the practice of feeding meat-and-bone meal to young calves of dairy cows.

    • fifi@piefed.zip
      link
      fedilink
      English
      arrow-up
      14
      ·
      1 day ago

      At the same time they are also pulling towards it because of the potential recursive self-improvement has. It’s not about just the training data anymore, it’s AI finding new training techniques, new kernel architecture and so on. It’s a separate, potential scaling vector as the training data vector has pretty much been maxed out in terms of big benefits.