AI art for TUGs?

[BDBrit]

I was recently exploring online and saw a picture that was AI generated, which got me thinking about whether AI could be used to illustrate stories on here. It looks like a lot of sites require you to sign up and pay for the privilege. I found a couple of free ones but they were pretty rubbish. The results ranged from blurry to outright horror shows. Safe to say the AI needs more training on what 'tied up' and 'gagged' mean.

I don't know what this forum's position is on AI art so I won't post any examples yet but I was wondering if anyone more knowledgeable than me might know a good website or app to try. I hesitate to pay for something that'll be rubbish.

[FelixSH]

I tried a bunch of sites some time ago. None could produce any form of bondage, be it a tied up person, or just a gag. Might be me - maybe, if I had given different prompts, I could have gotten something better. But it seems more like these ai generators were simply unable to (my guess is, that the creators simply don't want it to be used for bondage, because it's dirty and gross, or whatever nonsense many people still think about it).

But you could certainly use it to create portraits for characters, most AI generators can create pretty nice faces or even full bodies, by now. Often, if you have a whole body, there is something a little bit off, in my experience, but portraits seemed fine. Just don't let it create hands, they often came out terrifying.

Important with AI artwork is, that it is created by feeding the AI art that already exists. Tons of it, and I guess no one was asked if the generator was allowed to use it. Which means that you use parts of the artwork of others, without their consent and without paying them, if you use AI generated artwork. Which I'm not critizising, but one should understand that peoples work is, in some capacity, used without their consent.

I do think, that for personal, small things (like a story for this forum, where no one earns any money from), AI artwork is fine. Maybe I'm wrong, and you shouldn't use anything from AI, as soon as you do it publicly. But as soon as you earn money, you also should be ready to pay an actual artist, instead of using AI. In any case, if you use AI artwork, I feel you should say so. Acting like you created it, when it comes from someone or something else, just feels wrong to me. But, again, just my opinion.

Reading your post again, I feel like I got off topic here, this wasn't really your question. But this is a thread about AI art, so I feel like it's the right place for this discussion, so I'll just leave it here. I do feel like it would be good, if the mods and admins could adjust the rules, so it is clear if AI artwork is allowed here, or not.

[BDBrit]

Thank you so much for such a measured and thoughtful reply. It's given me a few interesting things to ponder. The internet can be a very depressing place a lot of the time but sometimes we are able to have fascinating conversations with strangers. I don't know a lot about this topic so you've shown me another way of thinking about it.

[Xtc]

Personal opinion: I don't see why AI generated crap (Oops! sorry; art) should not be posted here but I agree with everything that [mention]FelixSH[/mention] said above.

What do you think, [mention]chadmc90[/mention] and [mention]Johnsnow[/mention] ?

[Lara]

Everybody has their own definition of art. In my case "AI" generated data (whatever medium) are not art. I think something can only be considered art if it carries some form of message or intention to the recipients. "AI" generated works don't carry any message. They are just formed by superpositioning probability functions calculated by many computational nodes. It's still only machine learning, there is nothing intelligent about it.

Since such "AI"-images don't carry any meaning, I don't see the point to add them to the story. Usually with any good story, my own imagination is enough, there is absolutely no need for illustration. If you feel a story cannot be enjoyed without illustrations the story itself may be the problem and should be fixed first.

Unedited "AI"-images almost never make sense. The degree of sharpness meanders throughout the picture in a comical way, contradicting the usual physical principals of how a lens works. Catchlights in the eyes don't fit the lighting or look weird, the pupils and irises are often not round / oval but look like a square with rounded edges. Hair is unnatural or super blurry. Once you zoom in, you can sometimes see the background colliding with foreground objects like in a bad video game with no physics engine.

So if people publish such images, they should be able to deal with negative criticism if the images look too bad. Or be prepared to invest the time in manual photo retouching beforehand.

I suggest to make it transparent by a footnote, which images are "AI" generated. Same for stories: those threads with generated stories should be marked with a specific tag in the title or have their own subforum, so people who don't enjoy such "great" "AI" "art" can better ignore it.

[potato]

I personally think AI art is exactly that: Art.

(Buckle up, this is gonna be a long one. I added a TL;DR just in case somebody doesn't want to read everything. If you disagree with my point after only reading the TL;DR, scroll back up and see my argument in full detail)

Now, there are a few reasons why as to I think this, but let me propose a scenario:

Let's say you're an artist, and you're working with an author about this scene in the story you both are making. The author has a vague idea as to what they want for this scene, perhaps they know what the character's hair color is, or what they're wearing. But whatever the case, they only give you about a sentence or so worth of stuff to go off of.

So you go online to search for inspiration, perhaps you don't know what the author exactly meant by "schoolgirl outfit" or something. You look around multiple pieces of artwork from a variety of other talented artists. And you like what you see.

Perhaps this one design from this one artist looked really cool in your opinion, so you take inspiration and base your own design off of their design. Using what like liked most about it for your design, making sure to not plagiarize their original work.

In this case, this artist symbolizes AI (albeit to a smaller degree). There is no plagiarism, since by definition, plagiarism is "the practice of taking someone else's work or ideas and passing them off as one's own."
Inspiration, on the other hand, is defined as "the process of being mentally stimulated to do or feel something, especially to do something creative."

So plagiarism would be the AI directly taking an artist's artwork and presenting it to you as "AI-generated."
But inspiration would be basing ideas off of lots of artists and then molding it into something new and original.

And this is a big argument for the anti-AI side, is that AI art directly steals artworks from artists without permission.

Which, is totally false. You need inspiration and creativity to make art, and that's what makes art beautiful. Because you take that inspiration, and mold it into something new. And to say that AI can't take ideas from artists is basically telling artists to not use their creativity.

Let me "debunk" a couple more arguments left in the comments:

1. "Unedited AI images almost never make sense"
This left me a bit confused. To use AI not having a good first generation image as your supporting evidence as to why AI isn't art, is kind of weak. Just because something is not good, or has flaws, in no way means it is not art. If I draw something, and it has flaws, does that mean my own image that I draw with pen and paper isn't art? No, right? So then why should my flaws be counted towards the definition of "art" but for AI it is counted against it?

2. "AI art has no meaning"
Art does not need a meaning to be art. Sure, meaning adds a lot more depth to it. But if I were to ask you, what is the meaning of "The Plate of Apples" by Paul Cézanne. I would bet that you wouldn't know it. For those of you who would say "it captures the various viewpoints of a simple object" as since, that is pretty much what still life art is. I ask you this:

Now, if I were to ask AI to generate "A moment between a grandfather and grandchild" and it showed an old man and his grandson, sitting together on a small fishing boat. They have a look of laughter on their faces, their hands clutched onto a fishing pole.

Does that scene not have a meaning? Does it not carry any emotion to you?


I leave off with this statement: Feel free to disagree. Feel free to argue back. Feel free to prove me wrong. This is why this debate on AI art is exactly that, a debate. Each side has equally strong evidence for their point. If one side was clearly wrong, well, it would be like how they are somehow still flat-earthers all over the globe.

Again, I'm open to discussion. As each person's opinion needs to be heard.

TL;DR
- AI art does NOT plagiarize, it takes inspiration
- [Debunk]: Just because AI has flaws doesn't mean it's not art
- [Debunk]: AI art does have meaning, just because it wasn't drawn by a pen doesn't mean the image carries no emotion

[potato]

Just realized I forgot one other big argument.

This is especially prominent with human artists. They all say that they disapprove of AI art because it allows regular people to create artworks that took them years to be able to master. And that it's "unfair" for them, as they would have no use anymore.

And for me. this is perhaps one of the dumber reasons to hate on AI.

Let me propose this: If a lumberjack from the early 1800's, cutting down trees with nothing but an axe, saw a man use a chainsaw, how would he react? Would he be mad that some regular person can do his job just as well, but faster?

Sure. Or perhaps he wouldn't.

But what about you? Do you think creating something that could replace humans in order to be faster and more effective? Imagine all of the infrastructure that we wouldn't have if we were still stuck chopping down trees with our hands.

My point is: Creating new and powerful tools to be better than humans is simply the advancement of humanity. It might seem unfair to you, if you see your job being stripped away from you by a machine, but it's basically another step forward for humanity.

Take a look back at everything, what things have we created to be better than us?

Cars to replace carriages.
Telegram to replace letters.
Automated factories to replace hundreds of workers.

Each and every technological development created that propelled humans forward obviously had some impact on the things they were good at. And I see it as an unfortunate but unavoidable casualty.

[Lara]

Well, I don't hate AI tools, just have a different opinion about these. You're mixing some of my arguments with other peoples', making it difficult for me to respond because I don't know when you're addressing me and when it's a general remark. Anyway, I think we agree on AI being a potentially useful tool, and I too do not consider it plagiarism and am happy if AI removes some forms of gate-keeping issues.

I have not said AI is not art because it has flaws. I meant to talk about a practical problem: I assume, most people here expect some degree of quality for story illustrations and I'm not sure how useful images of square-eyed 6-fingered people melting into backgrounds are... Personally, so far, I can't enjoy AI works. No matter if we regard stories, images or music, here or someplace else. A number of people posted such works and only later revealed the fact that these had been created with an AI. Hence when I first viewed them, I did not know it, so we can consider this to be a blind test. In all cases I could not enjoy these works for reasons I wrote in my first post but also for other reasons. But then, if I can't enjoy AI works, what's the point? I think I've read through all ChatGPT stories that were posted here lately, and honestly they all suck. There was some core idea worth developing, but in the given form the stories were boring, clichéd, illogical and all characters were so dull. I'm not willing to invest more time for nothing and am going to ignore them now. Human writers do a better job on average. As do the human photographers or graphic artists or musicians. If generated works are neither interesting, nor emotionally moving, nor meaningful to me, I gladly ignore them.

Also, I think the reason why I can spot AI stuff so easily is that it was created unintentionally by a machine. In contrast, even artworks that do not carry any deeper meaning beyond being a artifact of the fleeting existance of humans, were still created intentionally, right? The painter who decided to hang an unpainted, virgin canvas in the gallery did make this decision with intention, right? An AI that decides to return nothing when prompted to paint a picture is probably considered to have run into an error state by pretty much all users - they would not think the AI has its own will and intentionally stopped working.

Finally, an AI can be a useful tool, and a dangerouos tool. Earlier predictions in regard to automation promised that everybody would need to work very little and be vastly wealthy. In reality fewer people than ever can even afford their own place to live despite working 40 hours and more a week. Instead we observe an incredible accumulation of wealth at the top. We need to be careful. AI systems will help us but also create unforeseen problems. It might be a good idea to make the companies running those AIs not too powerful. AIs run on energy-hungry cloud systems, someone has to pay the bill...eventually...

[puddy300]

One of the biggest problems communities are seeing from AI art (beyond the ethical issues) is that many of them who allow it become inundated with same-y AI art (since it's so easy for anyone to produce at high volumes) and drowns out anyone who puts actual effort into their work.

Also I fundamentally disagree with the idea that creative and artistic endeavors can be equated to drudgery work like working a factory.

[potato]

Firstly, I'd like to agree with you all on some points.

Yes, having a giant wealth gap is bad. However, these people who pioneered these inventions do deserve their wealth. I bet you that you're thankful for Larry Page and Sergey Brin for inventing Google. Should they pay less taxes then us? Probably not. But that's a completely different topic/debate.

And yes, the fact that anyone can produce a high quantity of images does pose a problem. Especially to human artists. I can see why that is a problem.

However, you mixed up my point with that last one. I never said that works of art is equal to factory work. The only reason I put my analogy up there is for that exact reason: an analogy.

The point I was trying to make there was that AI art is just physically better than human created art.
Now, before you rebut this claim, let me break down what I mean.

I just mean physically. People can argue that human-made art is so much more complex and meaningful than AI-art, but that's not my point. My point is that AI-art is just faster, more efficient, and more effective.

Now, notice in my arguments that I use humanity as a whole as my subject. This means as a whole. I'm not talking about the next "da Vinci" or "Picasso" but instead I'm talking about the average Joe.

If this average Joe wanted to illustrate something, he'd be generally be presented with 3 options:

Option 1: Draw it himself
Option 2: Hire an artist to draw it for him
Option 3: Have AI draw it for him

While Option 2 is usually the "go-to" for many average Joes, whether it be commissions or a professional setting, it is the worst pick. Choosing this option means to pay up, especially in a professional setting. It also means having to wait, whether it be hours, days, weeks, or even months.

Option 1 is less harsh on your wallet, but it takes time. If you're already an artist, who can draw, this is by far the best option. Especially so if you don't support AI generated art. And I applaud those people. This option means you don't spend up to hundreds (or even thousands) of dollars at a time, and you know exactly what you want and when you'll finish. (Unless you procrastinate, of course.)

However, we're talking about an average Joe here. And I'm sure a lot of people never made it past the "stickman" phase in their artistic journey. And for these people, Option 3 is ultimately the best choice.

Using AI, you get an image in just a few seconds or minutes, you don't spend money (it's free) to get the image you want and you can get an image that's exactly what you want. (If you don't like it, tell the AI to re-draw it, simple as that.)


And one last thing: I don't believe I mixed up your arguments for the general public [mention]Lara[/mention].

In your argument, you stated "Since such "AI"-images don't carry any meaning..." and "Unedited "AI"-images almost never make sense."

The first argument was talking about how AI images don't carry meaning and in a good story, your imagination is enough. Which I can agree with. I mean, look at the "Harry Potter" series. It's just words and yet it's a successful franchise. However, art does provide better visualization of the scene.

The second argument was criticizing AI-art's image rendering flaws. You pointed out how AI generated hair looks unnatural and how the pupils and eyes look square. But the same kind of criticism can be applied to human artists too.

I responded with (direct copy-paste from my previous comment):
- [Debunk]: Just because AI has flaws doesn't mean it's not art
- [Debunk]: AI art does have meaning, just because it wasn't drawn by a pen doesn't mean the image carries no emotion

So which points did I reply to that weren't the points you said in your argument from (at the time I'm posting this reply) 5 days ago?

[FelixSH]

Using AI, you get an image in just a few seconds or minutes, you don't spend money (it's free) to get the image you want and you can get an image that's exactly what you want. (If you don't like it, tell the AI to re-draw it, simple as that.)

It's not free. A lot of people put a lot of work into drawing pictures. That's the value that is added. The AI, which just takes from them, never gives anything back.

If you use art to earn money, pay the artist. Everything else is just greedy. Someone put the work in, so that the AI can use it and base it's output on it. Just pay someone from deviantart 100 bucks, or something. If you make money from your product, you probably will get it back. Or give them a share. You use their work.

I don't even care, if this is fair against AI. Just do it, because it's the right thing to do. You help an artist, who probably needs it anyway. The AI doesn't. Yeah, you spend a bit of money. So what? You get something out of it. Help an artist. The AI doesn't care. The artist will.

If you do it on your own, knock yourself out. Use AI as much as you like. But if you earn money? Do the right thing, and help an artist out. That you can save every last cent doesn't mean you should. To the artist, it probably means the world, to not only have their work recognized, but to also get money from it.

Also, show your work. If you use AI created stuff, mention that it's made by an AI. Be honest about it. I...let's say I don't think very highly of people, who act like they created something that they didn't (aka they stole art). Even if you don't get money out of it, it's just, dunno, kinda pathetic, lying for something like that.

An important point with art is that it's creative. It's expressive. It might be "bad", for whatever that word is worth anything, when talking about art, but there is someone trying to express themselves. Which can lead to very chaotic, horrifying, maybe disgusting stuff, but stuff that is at least interesting.

You will not get any of this from an AI. They use calculations. There is no creative spark. They will always just copy what is already there, because they can't do anything else. They can't add anything from themselves, because they don't have anything to add. And this is, I think, the main difference, why AI can't create art. At least not now. It might change, but the AI we have now is far away from anything really creative.

As long as it can't give me anything daring, anything weird and off-putting, anything hard and mentally chewy, I will not considere it art. In general. Humans add something to the art they create. AI doesn't. That's why I don't consider it art.

[blackbound]

Everything else aside, and I'm fully with FelixSH here, the publically available "AI" models are as milquetoast centrist as humanly possible and are definitely not trained on TUG images, so you likely won't get anything useful unless you spun up and trained your own instance.

[potato]

You brought up some really good points. And I'll agree with you there.

Supporting artists does seem like a good cause to reason to spend your money. And art thieves are definitely scum. (However, AI art is different compared to just straight theft.)

However, AI art is free.

Go to Stable Diffusion, see the prompt line, type in what you want, and you get an image.

My point was a purely "you don't spend money" standpoint. (I think you can call it "economical")

You don't pay a cent, you don't give your credit card info or e-mail, you don't Venmo or Cashapp someone.

It's free.

[FelixSH]

But it's not free. Artists already put effort in, elsewise the AI wouldn't have any access to the art it imitates. They paid for what the AI creates. Nothing is free, everything has a price. Just because we don't pay for it, when we use an AI, doesn't mean there isn't a cost. It's just paid by others.

Partly, it's the work of a ton of artists, who put effort into their work. These ARE costs. Don't just focus on the stuff the consumer pays. It's not everything.

You can call it economical. I call it not paying the people who did the work. That you can, doesn't mean you shouldn't. Don't pinch every penny. Give back to other people.

And yes, I still call it theft. The AI uses art, and uses functions and methods to mix that into something else. It's not new. There is nothing from the AI itself. It just uses stuff that was fed into it. And doesn't give credit to that stuff. It's theft. If this is unfair against the AI, I don't care. It's not a person. I don't need to give it the benefit of the doubt.

In the end, I only care about two things:
- Be honest, and say that you used AI to create something. Like with all other art, don't act like you created something you didn't.
- If you earn money from it, pay an actual artist. Even if AI created stuff isn't art (I actually don't care for that debate, sometimes I just fall into that hole). Even if it isn't theft. Just because it's the right thing. I think there is even art that is free to use, that doesn't look bad (need to confirm that, though), then use that and give credit. Help each other out. Simply because it's right. Don't save every single cent, just because you can.

Anyway, I'm getting out of this debate now. It doesn't really matter, at least here, and I made my point. For me, it's more about what is right, instead of debating single points, and to me, that means paying artists, if you can, instead of making sure you pay as little as humanly possible. If you earn money from something, like a game, you can. Give something back.

Anyway, I'm out.

[Lara]

However, you mixed up my point with that last one. I never said that works of art is equal to factory work. The only reason I put my analogy up there is for that exact reason: an analogy.

Sorry, this last "Finally, ..." paragraph was meant to sum up the disucssion on my end. Besides, I too can use an analogy, can't I? This argument is there to warn of being too blindly in favor of something. In fact neither you, nor I equated AI to factory work. My message was: don't trust promises that are too good. AI won't be the great technology that automatically revolutionizes everything in a good way. We need to be watchful and set rules to shape the usuage of AI in such a way, that we can really benefit from it in the end.

I just mean physically. People can argue that human-made art is so much more complex and meaningful than AI-art, but that's not my point. My point is that AI-art is just faster, more efficient, and more effective.

I think the statement is (at least for now) uncertain. Nobody knows yet if AI is more effective in the field of art or not. To my knowledge, there has not been a study proving that. For example, with a sample size of a few AI hundred images/stories, that I've seen by now, there was not a single one that I liked. The effectivness is therefore 0 in my case. The question is how many people feel the same? How many people can you reach with AI works and how many can be reached with traditional art? What's the point of being able to produce art quicklier, if you reach less people? Of course, it could turn out, that AI reaches more people in the end at some point in the future. Right now we just don't know. Also you seem to forget the costs of development, operation, energy cost, fixing security issues, updating components, etc. I've been working with Azure and AWS on my job. I know how much work there is. Of course you need to factor that in and the effort to create the trainingset and have the original artworks of the traingingset produced in the first place - from which the AI is entirely dependend. So if you still argue the AI is more efficient to produce a generated painting, this is at least to some degree unfair. Anyway, an artist who paints manually consumes far less energy. The cloud system, the internet, and an operation team are all required for every single AI generated image, whereas the painter just needs a canvas, some paint and a brush. The production costs (energy and resource wise) for millions of brushes, etc, seem to be lower than that of entire datacenters. So I really don't subscribe to your confidence of AI being more efficient and effective. Not until there are reliable studies taking all this into account.

While Option 2 is usually the "go-to" for many average Joes, whether it be commissions or a professional setting, it is the worst pick. Choosing this option means to pay up, especially in a professional setting. It also means having to wait, whether it be hours, days, weeks, or even months.

Option 1 is less harsh on your wallet, but it takes time. If you're already an artist, who can draw, this is by far the best option. Especially so if you don't support AI generated art. And I applaud those people. This option means you don't spend up to hundreds (or even thousands) of dollars at a time, and you know exactly what you want and when you'll finish. (Unless you procrastinate, of course.)

However, we're talking about an average Joe here. And I'm sure a lot of people never made it past the "stickman" phase in their artistic journey. And for these people, Option 3 is ultimately the best choice.

The best option for the average Joe, is exactly the option they choose given their individual circumstances.

It's important what they want to do with it, just hang it in their living room or publish it on social media? The latter is coming with legal risks. E.g. using AI comes with potential legal issues about copyright, that are unclear at the moment. You say the costs of option 2 is problematic, but at the same time ignore the potential costs when using options 3 - it could result in a loss of thousands of dollars if the person loses a lawsuit. You have to multiply the likelihood of a lawsuit coming up and losing it with the potential total costs and compare that to the costs of hiring. As we all know from the media there are people trying to proceed in court against AI companies and we need to wait for the outcome. Only after that we can estimate the probability of a copyright lawsuit. Maybe all those claims are turned down, or some of them are entitled to receive compensation, who knows. It's also unclear if the AI company or the user who publishes a work are liable or both. So if Joe wants to play safe, option 1 or 2 are clearly better than 3. But anyway, it's up to them and I think it's pretty difficult to predict which options they prefer. There are also many more options, like paying a friend to paint a picture by inviting them to a dinner. A colleague of mine did that.

And one last thing: I don't believe I mixed up your arguments for the general public @Lara.
So which points did I reply to that weren't the points you said in your argument from (at the time I'm posting this reply) 5 days ago?

Yeah, I see now, you probably addressed my arguments below "Let me "debunk" a couple more arguments left in the comments:" and all points above that are enitrely unrelated to my points, right? I wasn't sure. Sorry about that. But on the other hand, not sorry, because I felt your wording to be disrespectful and was a bit upset when typing my reply. You can use "debunk" when disproving a claim with scientific or objective facts. But when talking about meaning of art there is no clear right or wrong, so the use of "debunking" was not justified because you disvalued my opinion. And since we're on that issue: you wrote: "But if I were to ask you, what is the meaning of "The Plate of Apples" by Paul Cézanne. I would bet that you wouldn't know it." - I perceived this as condescending. And by chance you picked a bad example, because I do have quite some meaningful things attributed to that painting. Whether the artist had the same meaning in mind, does not matter to me.

----

Finally, as FelixSH also said, AI art is not free. As I mentioned earlier, it costs energy, somebody has to maintain the cloud system, fix security bugs, at some point introduce new works into a new training set, etc. I'm pretty sure free access of the prompt will be limited in the future. No company just offers free content without a catch and all this infrastructure costs them a lot. Oh, and [mention]potato[/mention], you forget the energy and internet costs and end-user devices. Devices don't grow in the woods, internet does not rain down, you have to pay for them (or someone). Anyway, you can't use AI without those. The cheapest is the library, but then the state or whoever finances the library pays the bills. We don't live in a world where things are free.

---

I'm out here too, said everything that is to say.

[tiedinbluetights]

Well said [mention]FelixSH[/mention]! Thanks for so eloquently pointing out the real issues with these so called AIs.

And also thanks [mention]Lara[/mention]! Your posts are thoughtful and meaningful, I only saw your reply after I posted the above line.

Me thinks I'll follow suit and no longer pay attention to this particular thread.

[potato]

Well said, all of you. I'm starting to see your points and you pointed out my argument's flaws.

I'll leave the argument off with that.

I'll go silent as well.

[Terry]

Tried to use Midjourney for this but it blocks words like "tied up"unfortunately.

[Monty]

However, you mixed up my point with that last one. I never said that works of art is equal to factory work. The only reason I put my analogy up there is for that exact reason: an analogy.

Sorry, this last "Finally, ..." paragraph was meant to sum up the disucssion on my end. Besides, I too can use an analogy, can't I? This argument is there to warn of being too blindly in favor of something. In fact neither you, nor I equated AI to factory work. My message was: don't trust promises that are too good. AI won't be the great technology that automatically revolutionizes everything in a good way. We need to be watchful and set rules to shape the usuage of AI in such a way, that we can really benefit from it in the end.

I just mean physically. People can argue that human-made art is so much more complex and meaningful than AI-art, but that's not my point. My point is that AI-art is just faster, more efficient, and more effective.

I think the statement is (at least for now) uncertain. Nobody knows yet if AI is more effective in the field of art or not. To my knowledge, there has not been a study proving that. For example, with a sample size of a few AI hundred images/stories, that I've seen by now, there was not a single one that I liked. The effectivness is therefore 0 in my case. The question is how many people feel the same? How many people can you reach with AI works and how many can be reached with traditional art? What's the point of being able to produce art quicklier, if you reach less people? Of course, it could turn out, that AI reaches more people in the end at some point in the future. Right now we just don't know. Also you seem to forget the costs of development, operation, energy cost, fixing security issues, updating components, etc. I've been working with Azure and AWS on my job. I know how much work there is. Of course you need to factor that in and the effort to create the trainingset and have the original artworks of the traingingset produced in the first place - from which the AI is entirely dependend. So if you still argue the AI is more efficient to produce a generated painting, this is at least to some degree unfair. Anyway, an artist who paints manually consumes far less energy. The cloud system, the internet, and an operation team are all required for every single AI generated image, whereas the painter just needs a canvas, some paint and a brush. The production costs (energy and resource wise) for millions of brushes, etc, seem to be lower than that of entire datacenters. So I really don't subscribe to your confidence of AI being more efficient and effective. Not until there are reliable studies taking all this into account.

While Option 2 is usually the "go-to" for many average Joes, whether it be commissions or a professional setting, it is the worst pick. Choosing this option means to pay up, especially in a professional setting. It also means having to wait, whether it be hours, days, weeks, or even months.

Option 1 is less harsh on your wallet, but it takes time. If you're already an artist, who can draw, this is by far the best option. Especially so if you don't support AI generated art. And I applaud those people. This option means you don't spend up to hundreds (or even thousands) of dollars at a time, and you know exactly what you want and when you'll finish. (Unless you procrastinate, of course.)

However, we're talking about an average Joe here. And I'm sure a lot of people never made it past the "stickman" phase in their artistic journey. And for these people, Option 3 is ultimately the best choice.

The best option for the average Joe, is exactly the option they choose given their individual circumstances.

It's important what they want to do with it, just hang it in their living room or publish it on social media? The latter is coming with legal risks. E.g. using AI comes with potential legal issues about copyright, that are unclear at the moment. You say the costs of option 2 is problematic, but at the same time ignore the potential costs when using options 3 - it could result in a loss of thousands of dollars if the person loses a lawsuit. You have to multiply the likelihood of a lawsuit coming up and losing it with the potential total costs and compare that to the costs of hiring. As we all know from the media there are people trying to proceed in court against AI companies and we need to wait for the outcome. Only after that we can estimate the probability of a copyright lawsuit. Maybe all those claims are turned down, or some of them are entitled to receive compensation, who knows. It's also unclear if the AI company or the user who publishes a work are liable or both. So if Joe wants to play safe, option 1 or 2 are clearly better than 3. But anyway, it's up to them and I think it's pretty difficult to predict which options they prefer. There are also many more options, like paying a friend to paint a picture by inviting them to a dinner. A colleague of mine did that.

And one last thing: I don't believe I mixed up your arguments for the general public @Lara.
So which points did I reply to that weren't the points you said in your argument from (at the time I'm posting this reply) 5 days ago?

Yeah, I see now, you probably addressed my arguments below "Let me "debunk" a couple more arguments left in the comments:" and all points above that are enitrely unrelated to my points, right? I wasn't sure. Sorry about that. But on the other hand, not sorry, because I felt your wording to be disrespectful and was a bit upset when typing my reply. You can use "debunk" when disproving a claim with scientific or objective facts. But when talking about meaning of art there is no clear right or wrong, so the use of "debunking" was not justified because you disvalued my opinion. And since we're on that issue: you wrote: "But if I were to ask you, what is the meaning of "The Plate of Apples" by Paul Cézanne. I would bet that you wouldn't know it." - I perceived this as condescending. And by chance you picked a bad example, because I do have quite some meaningful things attributed to that painting. Whether the artist had the same meaning in mind, does not matter to me.

----

Finally, as FelixSH also said, AI art is not free. As I mentioned earlier, it costs energy, somebody has to maintain the cloud system, fix security bugs, at some point introduce new works into a new training set, etc. I'm pretty sure free access of the prompt will be limited in the future. No company just offers free content without a catch and all this infrastructure costs them a lot. Oh, and @potato, you forget the energy and internet costs and end-user devices. Devices don't grow in the woods, internet does not rain down, you have to pay for them (or someone). Anyway, you can't use AI without those. The cheapest is the library, but then the state or whoever finances the library pays the bills. We don't live in a world where things are free.

---

I'm out here too, said everything that is to say.

When a post from a bondage fetish site has more to say about AI Art than Twitter and Instagram...

However, to add to this:
AI is directly anti expression. Art, no matter who or what is made, is made for purposes, from people, to signify love or passion for something. AI Art goes against this. I'd much rather see a post on DA from Jam-Orbital, Liganometry, or someone like that, then see an AI spit out what I want. Everything should have a price, that's how it works, labor should be compensated with paychecks, however kinky they may be, so, commission artists too if you want something specific (As you pointed out), or alternatively, draw yourself (as you also pointed out)

[drgoremd]

Everybody has their own definition of art. In my case "AI" generated data (whatever medium) are not art. I think something can only be considered art if it carries some form of message or intention to the recipients.

Wrong.

I can squirt ketchup on a napkin and call it art. I can take a picture of a puddle and call it art. I can encase a pair of worn out out work boots in a Lucite box and call it art. I can record myself singing a popular song in an off key and annoying manner and call it art.

Art is anything the "artist" declares it to be. Not only does art not need to have a meaning, their have been entire schools of art, like Dadaists, who have argued that very concept.

[drgoremd]

AI is directly anti expression. Art, no matter who or what is made, is made for purposes, from people, to signify love or passion for something. AI Art goes against this.

Huh? So if I pay an artist to draw a picture of my cat he cannot do so unless he has a love of cats or a passion for drawing cats?

Or can he just be someone who is good at drawing and is willing to accept money to draw my cat which will allow him to draw the things he is actually passionate about in the future?

[Volobond]

Wrong.

I can squirt ketchup on a napkin and call it art. I can take a picture of a puddle and call it art. I can encase a pair of worn out out work boots in a Lucite box and call it art. I can record myself singing a popular song in an off key and annoying manner and call it art.

Art is anything the "artist" declares it to be. Not only does art not need to have a meaning, their have been entire schools of art, like Dadaists, who have argued that very concept.

Just say you don't really care about art or artists or the people behind art at all and move on, dude. AI cannot create art because it has no consciousness or intention. Art doesn't have to MEAN something, but it requires someone actually making it. An AI is just following algorithmic procedure. All the while impacting the livelihoods of real artists because people like you don't give a whit for expression or compassion.

[drgoremd]

Just say you don't really care about art or artists or the people behind art at all and move on, dude. AI cannot create art because it has no consciousness or intention. Art doesn't have to MEAN something, but it requires someone actually making it. An AI is just following algorithmic procedure. All the while impacting the livelihoods of real artists because people like you don't give a whit for expression or compassion.

A pen cannot create art, nor a pencil, nor a paintbrush because those things have no consciousness or intention either. However, when a human picks them up and moves them in a certain manner, they create art. The computer is just another tool like the pen, the pencil or the paintbrush. It produces art on behalf of the human who tells it what to do.

As for the "real artists", if people value a computer generated image over whatever it is they produce, then their art really isn't worth much or anything. Which begs the question, what exactly makes them a "real artist"?

[Volobond]

A pen cannot create art, nor a pencil, nor a paintbrush because those things have no consciousness or intention either. However, when a human picks them up and moves them in a certain manner, they create art. The computer is just another tool like the pen, the pencil or the paintbrush. It produces art on behalf of the human who tells it what to do.

As for the "real artists", if people value a computer generated image over whatever it is they produce, then their art really isn't worth much or anything. Which begs the question, what exactly makes them a "real artist"?

AI art does not have human intention behind it. You can put in keywords to prompt it, but that is not intending to create art - it's putting mincemeat in a grinder and hoping sausage comes out. It cannot reframe or create, only arrange data into something it has been told looks like what you want. That's nothing at all like using a tool like digital art programs or physical paintbrushes.

And as for your second point, you show a willful miaunderstanding of what makes art valuable, which is the creativity behind it. In a soullessly monetary sense, AI art cuts out the middleman, but at the cost of morality and actual people. If you want art, learn how to make it yourself or pay a real artist, because AI will never be able to make it. You can keep defending it all you want, but it seems all the more likely that you simply don't care about people or would rather have something for personal gain than help others.

[DeeperThanRed]

Just say you don't really care about art or artists or the people behind art at all and move on, dude. AI cannot create art because it has no consciousness or intention. Art doesn't have to MEAN something, but it requires someone actually making it. An AI is just following algorithmic procedure. All the while impacting the livelihoods of real artists because people like you don't give a whit for expression or compassion.

A pen cannot create art, nor a pencil, nor a paintbrush because those things have no consciousness or intention either. However, when a human picks them up and moves them in a certain manner, they create art. The computer is just another tool like the pen, the pencil or the paintbrush. It produces art on behalf of the human who tells it what to do.

As for the "real artists", if people value a computer generated image over whatever it is they produce, then their art really isn't worth much or anything. Which begs the question, what exactly makes them a "real artist"?

I think your analogy is flawed. I'm not a good artist but I can just pick up a pen and draw something I like, be it a character in my head or a simple sketch of my cats. There is inspiration and reference but in the end, that piece of artwork would not exist without me, regardless of quality.

Anyone can type the same words to an AI engine and they'll get pictures completely independent from them to similar results. The human component is irrelevant. And the produced content would not exist if not for real artists' (no quotation marks are necessary) artwork that the AI program uses without consent or compensation. You don't tell a paintbrush what to do, you actively need to participate in the creative process, which is what makes the end product unique and yours.

In the end, whether AI-generated content is "real art" or not, it is without a doubt unethical in the same way tracing and plagiarism are, which should be bigger concerns in my opinion. As others said, I don't think it's wrong to utilize AI for personal use but it becomes a problem when you start going for commercial use and belittling artists who actually put time and effort into bettering themselves and improving at putting the ideas on their heads onto paper.

The fact that you use other people's validation as a measure of an artwork's quality speaks volumes about how you miss the point of artistic creation, by the way. You can't gatekeep inherent artistic value with numbers. And even if you could, I've yet to see an AI artist whose works are valued as much as real artists like say, van Gogh or Monet.