• TheOneCurly@lemmy.theonecurly.page
    link
    fedilink
    English
    arrow-up
    2
    ·
    10 months ago

    Agreed, the output of the model isn’t really the issue (other than providing evidence that those works are in the dataset), the inputs are. I would also like to see groups like wikipedia get something back from this too, OpenAI abused a ton of community funded resources and expects to sell that shit back to us as GPT subs.

    • Anamana@feddit.de
      link
      fedilink
      arrow-up
      2
      ·
      10 months ago

      I see your point. If you’re interested in the topic, Jaron Lanier published a book about micro-payments for datasets in 2013. It’s an interesting read.

      If you’re too lazy you can also ask ChatGPT for a summary of the main points tho ;)