4
Vent: I just read that a major AI image generator was trained using over 100 billion images scraped from the web.
Found this in a tech article from a few days ago, and it's wild how much data they used without asking. It makes me question the whole 'innovation' angle when it's built on that scale of unlicensed work. Does anyone else think this kind of data sourcing is a huge problem for the field?
3 comments
Log in to join the discussion
Log In3 Comments
barbara_jenkins6613d ago
Honestly, I used to shrug it off too. But then I saw an artist friend's whole portfolio style copied in some AI art contest. It wasn't just one picture, it was her whole look. That's not just using public stuff, that's taking someone's work to replace them. How is that fair?
3
Unlicensed work" is a strong way to put it for stuff already posted publicly online.
1