www.inverse.com: Back when generative AI first reared its ugly head, there was a panic that the average viewer wouldn’t be able to identify it. But now, years into the rise of OpenAI, Grok, Midjourney, and the myriad other AI agents, anyone who has spent time online can identify a specific “AI look”: weirdly flat lighting, nonsense text, and often a strange yellowish tinge.
6 comments:
The title of this article doesn’t seem it accurately represent its contents as it actively describes how companies with licensing rights to anime and manga have been actively requesting OpenAI and other generative AI companies to not use their licensed material in the training of their programs and instead trying to take a route of in small increments get various productions produced for live action or another form of media. The main focus for these new projects is also to try and get the original writers involved more to ensure their success and try to recreate certain successes of Netflix in particular (especially since so many other adaptations have not proven successful). What I wonder more about this phenomena is when companies explicitly ask for their material not to be used in training, are there ways to fight against its generation or implement “anti-generative” features into AI? Or are these companies just at the mercy of other companies allowing their material to be used for training and then running the risk of similar content being able to be created? Further, what compensation or recourse is there for these companies if their material has been used?
There's this one image of Hayao Miyasaki floating around right now, where he has his fingers pressed to his temples and is gripping his pencil in his hand with anxiety. That's how I'm feeling about Artificial Intelligence, especially anime-based AI. Studio Ghibli is one of the last 2-D animation companies left in the world, and seeing the colloquially termed "AI Slop" in its style makes me sad. AI slop as described in this article's title can be a lot of things. Images, video, even those bunnies on the tampoline is AI slop. A lot of the time it's exactly what it sounds like, slop. But sometimes, like in the Studio Ghibli case, it's straight-up copyright infringement. Or malicious, which I think this is kind of both. The article doesn't actually talk about AI Slop that much, but I think that it's causing major issues all around the board. I think copyright laws need to be put into place to avoid things like this happening. I don't know how this would work, but i think its necessary.
I’m very glad to see a lot of studios formally requesting that their work be excluded from AI training. Knowing this is and has been an option definitely changes my views of Western movie studios, which are not putting pressure on AI companies about their training materials. At the end of the day, it really isn’t up to the studios and up to the law, unfortunately. We needed legislation about AI years ago; we require it now more than ever. Not just in regard to protected artistic works, such as film, but to individual likeness as well. The fact that there are no legal protections against this yet is truly mindblowing to me, seeing as Metallica was coming after kids on Limewire with all the rage the legal system could bring, not long ago. I see AI as copyright infringement with the intent to sell, and truly hope we get some legal protection against it soon.
I’ve always found it crazy how normalized it is that AI is allowed to simply copy the hard work of many artists to make its mediocre “art” without giving them any compensation. I’ve heard many arguments in favor of AI “art” stating that all art is derivative of previous works but I feel that AI learning from artwork is different from an artist taking inspiration from existing pieces. For one when an artist takes inspiration they inevitably already have their own style so even if trying to imitate the style of another artist they usually end up with a unique blend of their style and the style that they were trying to imitate. AI, however, does not have its own style and as such it recreates styles rather than taking inspiration from them. Even without the intellectual property issues behind stealing the style of an artist, making only direct reproductions of the style means that it cannot further the medium of art. The slight differences in how each artist interprets a certain style helps to find new and interesting styles. But AI isn’t able to make artistic decisions of when to follow one style and when to deviate from it, it can only do as it was taught.
I am so happy to finally see these creators and companies fighting back against AI. From the very beginning, I recognized these issues and was immediately horrified by how disrespectful AI is to artists of all kinds. Seeing all the “Ghiblify-ied” photos and videos online made me ill, especially since I knew at the time about Hayao Miyazaki’s stance on AI and how blatantly horrible and disrespectful it was to be using AI to plagiarize his and many other artists’ work. I am also glad to see that Square Enix is under the company that is fighting back against this technology, as I am a fan of their games. I am aware that an employee at Square Enix recently said that they wanted to use AI for debugging and that there were a lot of layoffs. This was extremely disheartening, and I was very upset, but I believe that community backlash and now this will serve to stop this course of action. I don’t believe they were ever intending to use it for art, but even then, I’d prefer they don’t use it at all, and this gives me hope.
I will never forget the video I saw of Studio Ghibli legend, Hayao Miyasaki, outwardly proclaiming his disdain for a pitch presented to him of one of the first attempts of using generative AI art. Granted, AI art has progressed a lot since then, becoming less and less indistinguishable from real art, but his harsh words have stuck with me and many others. Even within a culture as respectful in professional settings as Japan, Miyasaki absolutely chewed out the mere idea of using generative AI in an artistic context. He states, “I strongly feel that this is an insult to life itself.” Reading this article, while it’s comforting to see that studios are indeed fighting back against generative AI, the problem at hand is truly the ease of access anyone who can browse the internet has to use generative AI art for whatever they please. Restrictions should be placed on generative AI companies, not necessarily the studios themselves.
Post a Comment