this post was submitted on 12 Oct 2023
88 points (100.0% liked)
Asklemmy
43898 readers
930 users here now
A loosely moderated place to ask open-ended questions
Search asklemmy 🔍
If your post meets the following criteria, it's welcome here!
- Open-ended question
- Not offensive: at this point, we do not have the bandwidth to moderate overtly political discussions. Assume best intent and be excellent to each other.
- Not regarding using or support for Lemmy: context, see the list of support communities and tools for finding communities below
- Not ad nauseam inducing: please make sure it is a question that would be new to most members
- An actual topic of discussion
Looking for support?
Looking for a community?
~Icon~ ~by~ ~@Double_A@discuss.tchncs.de~
founded 5 years ago
MODERATORS
Question: are you aware that most AI "art" is actually just cobbled together assets from human artists, often without permission from said artists?
Serious question here. Genuinely wondering if you're aware.
Edit: I might've come across as rude. This was not intended and I apologize if it was.
Edit 2:
2023-11-11 – Was looking through some past comments and thought I'd edit this one with something important: I was wrong. The comments below mine taught me some important facts about AI art that I was not aware. These kind users took the time to educate me, and I appreciate that. AI art generation is a lot more nuanced than I gave it credit for. Please don't be past Me. Past Me was kind of a dick for their original comment here, even if they didn't fully mean to be.
Hey I got no problem with referencing art, but one should at least give credit for works referenced. Also, the AI isn't actually sentient, so it's just a generative algorithm; it's not actually creatively making derivative works.
So piracy isn't I feel a 1:1 comparison. At least in most forms of software piracy, the pirates still leave in the credits/splashscreen/whatever.
AI as a tool is not what I object to; what I object to is merely using AI to lazily slap together content without actually having creativity be the onus behind it, and then calling it art on the same level of art that does.
In short, there has to be an artist for something to be art; and for there to be an artist, there must be creativity, which requires sentience. Whether that sentience is from a true AI or a human is irrelevant, as long as it is there. If there is no creativity behind the "art", then generative AI in this case is not a paintbrush, but an assembly box pumping out bobbles.
The human actually has to have an idea of what they want? It's not just pressing a button, wait a bit, and there we go?
Is that what you're saying?
Am I understanding you correctly?
So you didn't just say "Draw Stalin as Captain Ahab on the deck of the Pequod"? You actually had to utilize some creativity for how it would actually look?
Credits?
Ooh, you shouldn't have done that. Now I get to have fun. >:3
Okay... Draw me Garras from Mass Effect 2 playing Marco Polo with Goku from Dragonball Z, with Princess Peach from Mario & Samus Aran from Metroid doing fisticuffs in the background. Do it in an ukiyo-e style.
Ready? Go!!!
I do indeed see what you mean.
Fair enough. I can admit when I'm wrong. :)
...except when a person is doing it, they're doing their own thing to it. They take an idea or two and filter it through their own lens and stylise it
Think about it like this - when you do data scraping, you're still interpreting the results. You're looking at the data and going 'ok from this I can draw X and Y conclusions based on this and that'. AI art is like if we removed you from the process - we just shoved all the data into a black box and it goes ding "X is Y". If you asked it why that's so, it wouldn't be able to tell you. You can't see how it works so you have no idea if it's reasoning makes scientific sense. It would not be admissible in a paper.
...don't most people kinda agree you don't pirate from small artists where piracy is actually hurting them? There's like, honour along thieves when it comes to piracy, and this is stepping all over the little guy who's actually hurt by this just to get your grubby little hands on something you think you're entitled to
Yes, I had heard that. But isn't all art? None of us is truly original. We are borrowing all the time. Even the greatest ones borrow.
Except I'm a sentient being; these so-called "AI" programs aren't actually sentient. They have no self-awareness. It's just a stream of IF/THEN statements with no actual awareness of said art.
I feel that's the difference. I have nothing against non-human art as a principle. When these "A.I." programs actually gain self-awareness and then create art, then I will gladly consider it genuine art.
I really don't care if it's "genuine" art or not. If it looks cool, it looks cool.
I can respect that. Lol.
There is not a single if/else in a neural network. You are confusing it with decision trees that are used for classification
Could you please explain? I don't think I understand.
Isn't every neural network, even the one(s) in our brain, just complicated "If A, then B" statements. Even just
"Given Image 1, Image 2, and Image 3, generate Image 4 by mixing them together according to Criteria 1 and 2"
would be equivalent to saying
IF((Image1, Image2, Image3) AND (Criterion1, Criterion2)),
THEN(Image4)
, would it not? :/
Edit: A word.
No, what you describe is a basic decision tree. Let's say the simplest possible ML algorithm, but it is not used as is in practice anywhere. Usually you find "forests" of more complex trees, and they cannot be used for generation, but are very powerful for labeling or regression (eli5 predict some number).
Generative models are based on multiple transformations of images or sentences in extremely complex, nested chains of vector functions, that can extract relevant information (such as concepts, conceptual similarities, and so on).
In practice (eli5), input is transformed in a vector and passed to a complex chain of vector multiplications and simple mathematical transformations until you get an output that in the vast majority of cases is original, i.e. not present in the training data. Non original outputs are possible in case of few "issues" in the training dataset or training process (unless explicitly asked).
In our brain there are no if/else, but electrical signals modulated and transformed, which is conceptually more similar to the generative models than to a decision tree.
In practice however our brain works very differently than generative models
I'm gonna be honest: I'm still rather confused. While I do now understand that perhaps our brains work differently than typical neural networks (or at least generative neural networks?), I do not yet comprehend how. But your explanation is a starting point. Thanks for that.
In the easiest example of a neuron in a artificial neural network, you take an image, you multiply every pixel by some weight, and you apply a very simple non linear transformation at the end. Any transformation is fine, but usually they are pretty trivial. Then you mix and match these neurons to create a neural network. The more complex the task, the more additional operations are added.
In our brain, a neuron binds some neurotransmitters that trigger a electrical signal, this electrical signal is modulated and finally triggers the release of a certain quantity of certain neurotransmitters on the other extreme of the neuron. Detailed, quantitative mechanisms are still not known. These neurons are put together in an extremely complex neural network, details of which are still unknown.
Artificial neural network started as an extremely coarse simulation of real neural networks. Just toy models to explain the concept. Since then, they diverged, evolving in a direction completely unrelated to real neural network, becoming their own thing.
I would like to add that if it were the case, that generative image "AIs" were if/else statement, they could not run on graphics cards, that are optimized for the same raw matrix calculations repeated on a lot of variables. If it was just if/else statement, they wouldn't need to do all the vector calculations stuff.
Just like your brain neurons.
You're comparing different things. That's not a valid, good-faith comparison.
Conscience arises from a complex system. Just like generative data does - to a different degree.
Except I'm self-aware. I have my own identity. These AI programs aren't true AI. They aren't self-aware; they don't have a distinct identity. One could argue that's really the only thing that separates an artist from a box with gears in it.
You missed my point. I wasn't saying AI is self-aware.
Prove it. Really, I don't think they are either, but I don't think you can possibly understand what sentience is there enough to make an intellectual property distinction.
... Especially given this. There's not really very many IF/THEN statements at all. Think more like a stack of basic math operations that are tuned by way of calculus, until we can't even interpret what they're doing.
That's actually why GPUs and even more specialised chips work so well for neural networks. There's no branching of question of what operation goes next.
Generative art allows more people to communicate with others in ways they couldn't before, and to inspire and be inspired by others. The stuff people post online still requires creativity, curiosity, experimentation, and refinement. It also requires learning how to use new skills they may not have had to effectively use new tools that are rapidly evolving and improving to express themselves. Generative art is not a passive process, but an active one, where human artists get a chance to create something unique and meaningful.
Think of it like a camera that that can navigate the multidimensional latent space filled with concepts that can give rise to novel digital art. In the real world you can up, down, left, right, in or out, but in a latent space not only can you go those places, you can go to where Muppets meets impasto. Like a camera, sometimes none of the things you capture are made by you, but you still choose how it's captured and presented.
You have a lot in common with Charles Baudelaire, even though you're a hundred years apart from eachother.
I believe that generative art, warts and all, is a vital new form of art that is shaking things up, challenging preconceptions, and getting people angry - just like art should. Generative art introduces new ways to fail that no one is ready for, so if you see someone post some malformed monstrosity somewhere, cut them some slack, they're just learning. Remember there's another person on the other end of the internet that was excited to share with you.
I hope you're doing well. It's cool you were able to better communicate that way.
Your comment definitely made me think, and I appreciate that, but at the risk of displaying ignorance, but in hopes of learning, I have to ask: doesn't creating A.I. art just involved pressing "start" (effectively) on the human's part?
If so, then the only one actually "doing art" would be the A.I., and since the A.I. is not self-aware, it's not actually an artist, just software gears-in-a-box pumping out a thing.
I fully recognize what all I just said may be misconceptions, but that's why I intentionally said it. If it's wrong, I can learn; if it's right, you can learn. No insult intended here.
It's kind of a spectrum depending on what tool you're using and your level of commitment. I know with web based interfaces wit can be slow and cumbersome to iterate, but with open source models based on Stable Diffusion you get a lot of freedom. That's mostly what I base my knowledge off.
Here are some videos of what I mean:
https://youtu.be/-JQDtzSaAuA?t=97
https://youtu.be/1d_jns4W1cM
https://www.youtube.com/watch?v=HtbEuERXSqk
Here is an alternative Piped link(s):
https://piped.video/-JQDtzSaAuA?t=97
https://piped.video/1d_jns4W1cM
Piped is a privacy-respecting open-source alternative frontend to YouTube.
I'm open-source; check me out at GitHub.
I'll gladly consider it proper art if the human actually uses genuine creativity in creating it. I'm not some troglodyte. So I appreciate you giving me information!
I'll check out those videos. Thanks. :)
It's kinda like selfies and cameras. Not all photos taken with cameras are art, but you can make art with cameras.
You make a good point, which is why some art is considered really good only because it’s unique. A lot of art is decent but derivative.
Seethe harder.
Shove off, you rude person. I'm trying to be polite with a genuine question and here you are being a dick. Fuck off.
You're not being polite or asking a genuine question, you're spreading lies under the guise of asking questions.
A "lie" by definition is intentionally giving a falsehood knowing it's a falsehood. I may be wrong, and I am open to being proved wrong, but that does not make me a liar. So I say once again: shove off.
Yeah, well you said this:
Which isn't a question at all, just a statement of fact disguised as an opinion, and a false one at that. Your only "question" is if the OP knew, which is a shitty and manipulative shaming tactic at best
So don't give my any crap about how you're just asking questions or you "aren't lying about anything because I could change my mind"
I was genuinely just wondering if they were aware. There was no malice in my question. It was not intended as a shaming tactic at all. You're reading way too much into this, frankly.
Never said that's why it wasn't a lie. If you're gonna give me shit for something, at least do it accurately.
That being said, this is getting rather ridiculous, so I'm out. You have fun now.
I think people know, they just don't give a shit because they think they're entitled to have art and artists are arseholes for not giving them exactly what they want
Only in the same way every human work with abstract influences is cobbled together from Picassos.
Source: I know how AIs work. In complete detail, for some systems.