Post AusYUcF35DcsWzZ37A by shanecelis@mastodon.gamedev.place
 (DIR) More posts by shanecelis@mastodon.gamedev.place
 (DIR) Post #AuqzyKy6B8VO5n740u by j_bertolotti@mathstodon.xyz
       2025-06-06T08:50:06Z
       
       2 likes, 3 repeats
       
       "AI companies claim their tools couldn't exist without training on copyrighted material. It turns out, they could — it's just really hard. To prove it, AI researchers trained a new model that's less powerful but much more ethical. That's because the LLM's dataset uses only public domain and openly licensed material."tl;dr: If you use public domain data (i.e. you don't steal from authors and creators) you can train a LLM just as good as what was cutting edge a couple of years ago. What makes it difficult is curating the data, but once the data has been curated once, in principle everyone can use it without having to go through the painful part.So the whole "we have to violate copyright and steal intellectual property" is (as everybody already knew) total BS.https://www.engadget.com/ai/it-turns-out-you-can-train-ai-models-without-copyrighted-material-174016619.html?src=rss
       
 (DIR) Post #AusYUcF35DcsWzZ37A by shanecelis@mastodon.gamedev.place
       2025-06-06T09:07:35Z
       
       0 likes, 1 repeats
       
       @j_bertolotti It'd be interesting if AI companies lobbied to increase the works in the public domain by decreasing copyright duration. That's something I'd actually support. Copyright is too long. And it would then be a legal, more ethical, industry instead of pack of VC-funded thieves. Strange bedfellows.
       
 (DIR) Post #Ausa6K6z2SPXGBKtou by lanodan@queer.hacktivis.me
       2025-06-07T06:59:15.049808Z
       
       0 likes, 0 repeats
       
       @shanecelis @j_bertolotti Thankfully they never would, imagine Microsoft waiving away parts of Windows code.And you can entirely have licences that allow to expire into a more permissive license or a public-domain equivalent, no need to screw around with everyone's copyrights.