I think this is completely missing the point when it’s talking about “the minutiae of art”. It’s making two claims at the same time: art is better when you suffer for it and the art is good whether or not you suffered. But none of that is relevant.
When Wyeth made Christina’s World, I don’t know if he suffered or not when painting that grass. What I do know is that he was a human with limited time and the fact that he spent so much of his time detailing every blade of grass means that he’s saying something. That The Oatmeal doesn’t draw backgrounds might be because he’s lazy, but he also doesn’t need them. These are choices we make to put effort in one part and ignore some other part.
AI doesn’t make choices. It doesn’t need to. A detailed background is exactly the same amount of work as a plain one. And so a generated picture has this evenly distributed level of detail, no focus at all. You don’t really know where to look, what’s important, what the picture is trying to say. Because it’s not saying anything. It isn’t a rat with a big butt, it’s just a cloud of noise that happens to resemble a rat with a big butt.
it’s just a cloud of noise that happens to resemble a rat with a big butt.
I’d like this on my tombstone
What a beautiful read. I feel the same about AI art and I remember a longer talk I had with my tattoo artist: ‘I need the money so I will do AI based tattoos my clients bring to me. But they have no soul, no story, no individuality. They are not a part of you.’
I feel the same.
Also I like Oatmeal’s reference to Wabi Sabi: The perfection of imperfection in every piece of art.
I often hear AI enthusiasts say that AI democratized art. As if art weren’t already democratized. Most anyone can pick up a pen, draw, write, type, move a mouse, etc. What AI democratizes in art, is the perception of skill. Which is why when you find out a piece of art was madeby inputting some short prompt into a generator, you brcomr disappointed. Because it would be cool, if the person actually had the skill to draw that. Pushing a few buttons to get that, not so much.
I have always felt that I’m not good at art (the practice I did got me not very far), and I’ve recently had reason to make little collages. One thing that I’ve done is uploaded pictures to Canva and traced them so I had something resembling recognizable images (my dog, me in a kayak). I don’t think tracing is making an art, AI is definitely not making an art.
Tracing is absolutely art. You choose what to trace what parts of the image are important what to discard etc.
“Yes, but I’ll be quick, I promise.”
Isn’t quick.
I appreciate this bit out of context:
Also loved the shoutout to Allie Brosh!
As a passable quality 3D artist who does it for a living I’ve found AI art (which can do 3D now to some degree) has kind of narrowed the scope for me. If you want generic Unreal style pseudo-realism or disney toon then AI can do that for you* I’ve had to focus much more on creating a unique style and also optimizing my work in ways that AI just doesn’t have the ability to do because they require longer chains of actual reasoning.
For AI in general I think this pattern holds, it can quickly create something generic and increasingly do it without extranious fingers but no matter how much you tweak a prompt its damn near impossible to get a specific idea into image form. Its like a hero shooter with skins VS actually creating your own character.
*Right now AI models use more tris to re-create the default blender cube than my entire lifetime portfolio but I’m assuming that can be resolved since we already have partially automated re-topology tools.
I forgot how loooong Oatmeal cartoons are. I don’t think I have made it to the end of one in years.
Wanna go ride a bicycle?
I made a comment about a week ago about how copying people’s art is still art, and it was a bit of an aha moment as I pinpointed for myself a big part of why I find image generators and the like so soulless, inwardly echoing a lot of what Inman lays out here.
All human made art, from the worst to the best, embodies the effort of the artist. Their intent and their skill. Their attempt to make something, to communicate something. It has meaning. All generative art does is barf up random noise that looks like pictures. It’s impressive technology, and I understand that it’s exciting, but it’s not art. If humans ever end up creating actual artificial intelligence, then we can talk about machine made art. Until then, it’s hardly more than a printer in terms of artistic merit.
There was a good interview with Tim Minchin by the BBC where he said something similar to this & used the word intent.
I suppose the intent/communication/art comes from the person writing the prompt but those few words can only convey so much information. When the choice of medium & every line etc. involves millions of micro-decisions by the artist there is so much more information encoded. Even if its copy & pasted bits of memes.Is this the interview? https://files.catbox.moe/ddp6tp.mp4
Tim Minchin has always come across as a good egg to me. It’s nice to hear he’s of the same mind, and I particularly like the optimism he’s promoting in his predictions for artistry going forward.
No, I hadn’t seen that one, thanks!
Hah, sure thing. I suppose it’s a point he’s discussed a few times.
It’s impressive technology, and I understand that it’s exciting, but it’s not art.
I would add that a lot (most?) graphical elements we encounter in daily lives do not require art or soul in the least. Stock images on web pages, logos, icons etc. are examples of graphical elements that are IMO perfectly fine to use AI image generation for. It’s the menial labour of the artist profession that is now being affected by modern automation much like so many other professions have been before them. All of them resisted so of course artists resist too.
The most generic logo from ten years ago still was made with choices by a designer. It’s those choices that make a difference, you don’t choose how things are executed with ai
But you still choose the final result…for something like that, the how is really quite irrelevant, it is just the end result that matters and that still remains in the hands of humans as they’re the ones to settle on the final solution.
the how is really quite irrelevant
That’s our point. The how is entirely relevant. It’s what makes art interesting and meaningful. Without the how and why, it’s just colors and noise.
I’d argue that logos are a hugely expressive form. It’s just that 90% of them are basic ass shit tier stuff.
AI has basically raised the level of “shit tier” pretty high. I sometimes go check out Hotone Audio’s Facebook page to see if there are new firmware updates for my device, but they mainly peddle pointless AI slop marketing images. I’m sure there are tons of companies like this.
It’s the literal example of the marketing person being able to churn out pictures without an artist being involved, and thus the output is a pile of crap even more vapid than stock photos.
The impact on livelihoods is important, but it’s ultimately unrelated to defining what art is. My consideration of art is not one born of fear of losing money, but purely out of appreciation for the craft. I don’t think it’s entirely fair to suggest all the criticisms against generated art is solely borne of self-preservation.
In regards to corporate “art”, all the things you listed, even stock images, are certainly not the purest form of artistry, but they still have (or, at least had) intent suffusing their creation. I suppose the question then is - is there a noticeable difference between the two for corporations? Will a generated logo have the same impact as a purposefully crafted on does? In my experience, the generated products I’ve noticed feel distinctly hollow. While past corporate assets are typically hollow shells of real art, generated assets are even less. They’re a pure concentration of corporate greed and demand, without the “bothersome” human element. Maybe that won’t matter in their course of business, but I think it might. Time will tell.
I was kinda against their argument at first, then I was with them and continued reading. But then they went into all sorts of detail, weighing pros and cons etc., and after reading more than half I evtl. gave up.
It seems all “why AI is bad” articles seem to go this way.
It seems all “why AI is bad” articles unwillingly even support the hype.
Fuck AI “art”, it’s not art you morons, it’s automation, which takes away real people’s jobs. The current implementations made by greedy companies also very obviously steal. 'nuff said.
I know that art is an art of it’s own and a way to express human creativity.
However people also complained once the loom was invented. It took lots of jobs.
The job argument is usually a stupid one.
The lack of creativity and quality is of course a much better argument against AI art.
The job argument is usually a stupid one.
The what? It’s the only one that objectively makes sense.
Ok imagine this:
You are an construction worker. The job is hard but the pay is okay.
Now robots replace your job slowly. They are cheaper and more accurate.
You can now:
-
Complain about the robots stealing your job
-
Be happy that you don’t have to do the hard work anymore.
Many people will go for 1. But the actual issue is that the social security net isn’t existent or so weak that no job means no food.
That is not the fault of technology though.
Remember that when you vote and when politicians want to cut costs by reducing payments for the unemployed.
Option 2 is soulless.
Option 3. Destroy the capitalists owned robots and bring the robots under the control of the working class.
Option 3 still ends up with robots and no-one doing the jobs that the robots replaced.
Option 3 would be a weird way of communism. Which still enforces my point. The reason why you fear for job safety is not the fault of technology.
-
It was an ok read for me, but mostly because I enjoyed the art rather than relating to the entirety of the sentiment.
I’m an artist and I find AI art evocative and illustrating things in a way that I wish that I could illustrate, but feel that is only because it comes from real human artists. I agree that it is a void in terms of difficulty to process, but there is still skill involved in both using search engines and describing something to an llm. A minute amount of skill, but still a skill.
I hate AI art because it is stealing from artists, not because it doesn’t feel right. It can have a million iterations and only needs to get it right once to count as feeling right to me. The relationship between the content and their artists to the ultimate product is removed, this to me is the wrongfulness of claiming new art from it. It is just stealing in a more wind-about manor. This isn’t like generating fractal art or something.
After all these years of corporations fucking up the literal social fabric and and how we communicate over IP law, for them to turn around and steal everything and just get a pass is an extra slap in face. Stealing only gets allowed2 one way in our society, and AI is just another example of that.
I’m honestly surprised to not see this take more from others and felt like i needed to mention it.
edit: emphasized that by making AI art taking skill, I only mean just a minute amount.
Thanks for sharing! I haven’t read much of the Oatmeal in quite a while but I’ve always liked their style and humor.
That was a beautiful read.
But do i find myself conflicted about dismissing it as a potential technical skill all together.
I have seen comfy-ui workflows that are build in a very complex way, some have the canvas devided in different zones, each having its own prompts. Some have no prompts and extract concepts like composition or color values from other files.
I compare these with collage-art which also exists from pre existing material to create something new.
Such tools take practice, there are choices to be made, there is a creative process but its mostly technological knowledge so if its about such it would be right to call it a technical skill.
The sad reality however, is how easy it is to remove parts of that complexity “because its to hard” and barebones it to simple prompt to output. At which point all technical skill fades and it becomes no different from the online generators you find.
All of that’s great and everything, but at the end of the day all of the commercial VLM art generators are trained on stolen art. That includes most of the VLMs that comfyui uses as a backend. They have their own cloud service now, that ties in with all the usual suspects.
So even if it has some potentially genuine artistic uses I have zero interest in using a commercial entity in any way to ‘generate’ art that they’ve taken elements for from artwork they stole from real artists. Its amoral.
If it’s all running locally on open source VLMs trained only on public data, then maybe - but that’s what… a tiny, tiny fraction of AI art? In the meantime I’m happy to dismiss it altogether as Ai slop.
How is that any different from “stealing” art in a collage, though? While courts have disagreed on the subject (in particular there’s a big difference between visual collage and music sampling with the latter being very restricted) there is a clear argument to be made that collage is a fair use of the original works, because the result is completely different.
Collage art retains the original components of the art, adding layers the viewer can explore and seek the source of, if desired.
VLMs on the other hand intentionally obscure the original works by sending them through filters and computer vision transformations to make the original work difficult to backtrace. This is no accident, its designed obfuscation.
The difference is intent - VLMs literally steal copies of art to generate their work for cynical tech bros. Classical collages take existing art and show it in a new light, with no intent to pass off the original source materials as their own creations.
only a note: LLMs are for text
Thanks. I edited
If you download a checkpoint from non trustworthy sources definitely and that is the majority of people, but also the majority that does not use the technical tools that deep nor cares about actual art (mostly porn if the largest distributor of models civitai is a reference).
The technical tool that allow actual creativity is called comfyui, and this is open source. I have yet to see anything that is even comparable. Other creative tools (like the krita plugin) use it as a backend.
I am willing to believe that someone with a soul for art and complex flows would also make their own models, which naturally allows much more creativity and is not that hard to do.
I think there’s a stark difference between crafting your own comfyui workflow, getting the right nodes and control nets and checkpoints and whatever, tweaking it until you get what you want, and someone telling an AI “make me a picture/video of X.”
The least AI-looking AI art is the kind that someone took effort to make their own. Just like any other tool.
Unfortunately, gen AI is a tool that gives relatively good results without any skill at all. So most people won’t bother to do the work to make it their own.
I think that, like nearly everything in life, there is nuance to this. But at the same time, we aren’t ready for the nuance because we’re being drowned by slop and it’s horrible.
I watched a short saying you might be an art director, at best, but not really an artist. Because you have the vision but you’re only telling someone (something) to materialize it. I was kind of happy with that.
Vision is a strong word. I think it’s a vague idea in most cases
I want to touch on how he mentions hitting the button to automatically make music on a Casio keyboard.
I fully realize I’m being reductive to the point of being offensive but that’s not my intent and I preemptively apologize, when I say: that’s at least in part, the very first seed to becoming a professional DJ. That’s not nothing.
Using AI to generate images can be the same thing if it’s extrapolated out into complexity and layered nuance. It might not make you an artist exactly, in the same way that a DJ might not be a musician but it IS a skillset that potentially has value.
And even if you think I’m totally off-base in saying so? I liked pretending with the little automatic music button on the keyboard.
Are you speaking from experience? 'Cause that’s not even vaguely related to how any of the DJs I know (including a couple of professionals) got started. The prime motive for most DJs is sharing cool music, and Casio keyboards don’t do that…
If I recall correctly it was a part of how Feel Good Inc. was made. That is tangential at best, I realize, but still a fun story.
No. Not from experience at all. I saw a small documentary once saying DJs remixed and sometimes create almost entirely new music and that they used computer based audio tools to do so. I’m probably thinking of a different profession. My ignorance, sorry.
All good, was just wondering.
I do DJ (non-professionally). I generally think there are two skills with DJing:
- Taste, library management and music choice, which is not a technical skill, but does take a bit of effort in preparing for a set
- Actual technical mixing skills, which many DJs (including me) barely have, but some take to a level that is on basically a form of musicianship.
I don’t think AI can really help you do either… but I guess it could make a mixed set and you could pretend to play it, like a Casio keyboard
I think pushing the button on a Casio keyboard is more akin to tracing your favorite comics panel than using an LLM image generator.
I mean…I liked doing that too. DIY Spider man coloring book pages.
This was a great read! As someone who was initially excited about the possibilities of AI art, it’s been hit or miss with me.
I’ve come to realise over time that I like the connection that art offers. The little moment of ‘I wonder what the artist was thinking when they imagined this and what experiences did someone have to get to a place where they could visualize and create this?’
And I think that’s what missing with AI art. Sure, it can enable someone like me who has no skill with drawing to create something but it doesn’t get to the point of putting my actual imagination down. The repeated tries can only get to point of ‘close enough’.
For me, looking at a piece and then learning it’s AI art is basically realizing that I’m looking at a computer generated imitation of someone’s imagination. Except the imitation was created by describing the art instead of the imitator ever looking at it. An connection I could have felt with original human is watered down as to be non-existent.
