DeviantArt upsets artists with its new AI art generator, DreamUp

The big problem here is that it trains the artists style and allows third parties to create art in their style without effort. This is especially bad for artists with really distinct style as now hundreds of copycats can come and steal the previously unique style.

Hate to inform you that the "cat is literally out of the bag" now. There's no putting it back. If it isn't DA it would be someone else. Right now, anyone who does anything creative like painters, authors, composers, designers have to live with the fact that AI can generate something similar to what they can and (this is the most tragic part) generate material that would be indistinguishable (or perhaps even superior) to the general public who're not connoisseurs of their art.

> This is especially bad for artists with really distinct style as now hundreds of copycats can come and steal the previously unique style.

Are you talking about human copycats? Same thing applied before AI models, to be honest.

I wonder what would Beeple say..

"Welp, I made my worth, so long suckers!"?

As opposed to copycats before the AI?.. Frankly, neither "stealing" nor "unique" make sense to me. Art styles aren't copyrightable for this exact reason - the entire culture is built on iterative variations, that's literally how it evolves.

I have to agree, this is a pointless battle. It doesn't matter what DA does or doesn't do. If your art is on the internet, it will be used to train AI. It's not even that it's an inevitable future, it's already the past. I do feel sorry for digital artists, but Pandora's box is open; A black ball has been drawn from the urn (in the world of digital art) and it's too late to do anything meaningful about it.

Copycating before required some effort (or as the techbros pedalling ai art might better understand a more familiar term, "proof-of-work").

A black ball for artists who don't use AI, for sure. The artists who jump on this trend will be benefiting the most as they know how to guide the AI better and can fix errors manually.

It still does, and will require effort and actual artistic skill and vision, due to entirely fundamental reasons (not because of deficiencies of current models). A lot of people who cry foul about AI art haven't actually dived into it, or thought at least a bit about what works and what doesn't. Typically they think that you can just enter a prompt and magically produce "art". It doesn't work like that, it's much more complicated, and the complexity is only going to increase in future. Just like with any CGI. This panic is induced by social media, and is based on a wrong premise.