• Auzy@beehaw.org
    link
    fedilink
    arrow-up
    7
    arrow-down
    2
    ·
    edit-2
    10 months ago

    I agree. It’s basically just stealing work from others and not paying them or agreeing to licensing. Long term, it’s not a good thing

    For software development, AI is a useful tool, but it’s likely also stealing licensed code and adding it to other people’s code in some circumstances

    If AI companies want to use training data, that’s ok, but they should pay all of the creators it’s trained from

    • NovaPrime@lemmy.ml
      link
      fedilink
      arrow-up
      5
      ·
      10 months ago

      Respect your opinion and where you’re coming from, but disagree simply for the fact that this will create a silo of corporatized AI because they will be the only ones who will have the ability to pay for the IP at the necessary scale (that’s not to say that this is not already happening to an extent with the existing model). I do think the conversation is worth having though about the public value of data that’s readily available on the internet and how it squares with our (imo) outdated IP laws. How do we ensure that individual creators retain full control and benefit of their art/content/knowledge, while not stifling or unduly hampering AI research? How much protection do we afford data that users willingly put on the internet that’s publicly available? And who pays for the data in the chain?

    • BlameThePeacock@lemmy.ca
      link
      fedilink
      English
      arrow-up
      4
      arrow-down
      1
      ·
      10 months ago

      Should human programmers have to pay creators every time they look at code while they’re learning? I (and most people) have literally copy/pasted code from various websites into my own programs without any sort of payment or notice.