I understand it can look uncanny. The fact that it was generated also changes perception at least a bit. On this one pic wings look strange. Because they are this long I expect them to be wider. Chicks look fuzzy when I zoom in to take a closer look. Though idk if I would have noticed these things if it wasn’t tagged.
It’s interesting to see how gen AI progressed compared to the previous AI pic posted here. IIRC that one was the first one and it was posted by me as a reason to discuss what should our stance on AI generated content be. Rule that we have right now didn’t exist yet.
I honestly don’t know what to think about gen AI. I don’t have negative opinion to completely forbid it in this community. Tag seems like a middle ground between completely accepting and banning AI. But if peoples think otherwise it would be cool if they responded why they think like this.
Comment on [AI] Niwatari Kutaka
hypertown@ani.social 1 week agoCurrently you don’t have a rule against this type of content so removing it would be kinda unfair… That being said you should ask yourself whether you’re fine with it. Personally I don’t want to see it, especially those uncanny ones.
vbb@lemmy.world 1 week ago
hypertown@ani.social 1 week ago
Forgive for being blunt but this pic is currently an example of one of the worst AI image generations. It’s very easy to spot even without zooming in. Currently there are models that are doing such a good job it’s hard to tell even after examining the fine details.
So I should be fine with the good ones right? After all if you can tell the difference then then it should be treated like a real art? … No! Even though I have to admit that some of them are really stunning visually I’m still very much against it. Why? Because I know how it works and that it’s highly unethical.
Image generation models are trained on art made by humans, this is probably a common knowledge, humans also learn from others but the difference is that human can create it’s own artstyle. AI usually copy some existing artstyle and creates a “drawing” based on that style. It’s very important to mention that this happens without the author’s permission! Not only that, many artists are very against AI training on their art but people do it anyway. This creates a very unpleasant reality where you spent years and years learning to draw, making your own style only to get that style stolen by machine that now can generate anything in a few seconds.
Someone might say: “So what? It’s called progress!”
No it’s not. This machine is nothing without artists and it is actively trying to put them out of their job.
Until AI (Or rather AGI) can create its own unique style not based on anything made before I’m going to hate it. But even if one day AI can create ethically something perfectly made for humans to enjoy I still don’t know if I want to consume things without a soul.Not to say I hate AI in general. I think it’s a very cool technology but it should be used as a tool to help people not to replace them.
So yeah, I don’t want to see it because if it’s bad it’s ugly and soulless and if it’s good it’s just some artist’s work reproduced without permission.
Small note that I used to post AI “art” too but that was before I knew how it works. I used to think it’s ok if people put work to improve and retouch the generated image. But the more I learn the more I’m against it.
If you really want to post AI slop then there are places that focus exclusively on AI. I use 3 different frontends for lemmy and not all of them support filters. If AI would flood this sub I’m just unsubscribing.
Even_Adder@lemmy.dbzer0.com 1 week ago
Community post tags are being added soon to lemmy so you won’t have to.
Lemmy works because you’re able to tailor your own experience, rather than trying to force your content preferences on everyone else. The way you carry on is unnecessarily divisive and tribalistic, and is going to cause lemmy to eat itself alive.
hypertown@ani.social 1 week ago
Tags are cool but they are not here yet and more importantly Lemmy is also not a place where anything goes. It’s up to admins and community to decide what’s allowed and what’s not. If the majority of the community doesn’t want something you can’t expect them to create tags just to exclude that something.
As a member of this community I’m just voting against AI mostly because it’s unethical, among other reasons.
Even_Adder@lemmy.dbzer0.com 1 week ago
Using things “without permission” forms the bedrock on which artistic expression and free speech are built upon. They want you to believe that analyzing things without permission somehow goes against copyright, when in reality, fair use is a part of copyright law, and the reason our discourse isn’t wholly controlled by mega-corporations and the rich.
What some people want will cripple essential resources like reviews, research, reverse engineering, and indexing information, and give mega-corps a monopoly of AI by making it prohibitively difficult for anyone else.
I recommend reading this article by Kit Walsh, and this one by Tory Noble staff attorneys at the EFF, this one by Katherine Klosek, the director of information policy and federal relations at the Association of Research Libraries, and these two by Cory Doctorow.
hypertown@ani.social 1 week ago
Copyright, it’s broken anyway. Some people abuse it, some are getting ripped off despite it. Fair use also is a concept that differs around the world. So I don’t know how legal AI training is and frankly I don’t care. All I care about is respecting authors. If they’re saying “No AI training” then you should respect it. This is also nothing like the Nintendo saying “Don’t reverse-engineer our products”.
Let me explain. Aside from the fact that you shouldn’t compare a soulless corporation to a living, breathing individual the most important difference is that for example Nintendo sells you a product in a way that even though you bought it you don’t own it. It’s your console so you should be able to reverse engineer, modify however you want it. But that doesn’t mean you can distribute their assets around the internet. You can’t just take Mario’s character model and put it into your game. For example every emulation project works like that.Using things without permission is fine as long as you won’t blatantly copy it and AI basically does exactly that. No respectable artists would do that, there’s no artistic expression in that. You’re just copying things. AI can’t get inspired by some work, can’t come up with a new personal artstyle. Maybe AGI can but AGI doesn’t exists yet.
I said this before but I’m going to say it again: This machine is nothing without artists and it tries to put them out of their work.
Besides there’s so much amazing art posted every day you won’t be able to keep up with it anyway. AI is not needed.
Even_Adder@lemmy.dbzer0.com 1 week ago
Please actually read the things I linked, they’ll explain this better than I can. Here are a few quotes:
Pluralistic: AI “art” and uncanniness
counting words and measuring pixels are not activities that you should need permission to perform, with or without a computer, even if the person whose words or pixels you’re counting doesn’t want you to. You should be able to look as hard as you want at the pixels in Kate Middleton’s family photos, or track the rise and fall of the Oxford comma, and you shouldn’t need anyone’s permission to do so.
How We Think About Copyright and AI Art
Moreover, AI systems learn to imitate a style not just from a single artist’s work, but from other human creations that are tagged as being “in the style of” another artist. Much of the information contributing to the AI’s imitation of style originates with images by other artists who are enjoying the freedom copyright law affords them to imitate a style without being considered a derivative work.
The people who train these systems still have rights like you and I, and the public interest transcends individual consent. Rights holders, even when they are living, breathing individuals, would always prefer to restrict our access to materials, but from an ethical standpoint, the benefits we see from of fair use and library lending, outweigh author permissions. We need to uphold a higher ethical standard here for the benefit of society so that we don’t end up building a utopia for corporations, bullies, and every wannabe autocrat, destroying open dialogue in the process.
What do you think someone who thinks you’re going to write an unfavorable review would say when you ask them permission to analyze their work? They’ll say no. One point for the scammers. When you ask someone to scrutinize their interactions online, what will they say? They’ll say no, one point for the misinformation spreaders. When you ask someone to analyze their thing for reverse engineering, what will they say? They’ll say no, one point for the monopolists. When you ask someone to analyze their data for indexing, what will they say? They’ll say no, one point for the obstructors.
And again, I urge you to read this article by Kit Walsh, and this one by Tory Noble, both of them staff attorneys at the EFF, this open letter by Katherine Klosek, the director of information policy and federal relations at the Association of Research Libraries, and these two blog posts by Cory Doctorow.
Even_Adder@lemmy.dbzer0.com 1 week ago
The reason they’re tagged is so you can filter them out. You don’t have to see them if you don’t want to.
remon@ani.social 1 week ago
The default lemmy UI doesn’t have any filter options, so people still do have to see them.
Even_Adder@lemmy.dbzer0.com 1 week ago
Lemmy.world has the Tesseract front end and community post tags are being added soon to the default one.