Lit really should reconsider the AI image policy.

[whitedudenj]

Here's one I made earlier. Norman Rockwell never dreamed.

Nice!
Got any more like that?

 

[Melody_Lightdraft2]

Using AI to generate images is just a major thing now whether we like it or not, this genie is not going back in the bottle and so Lit may as well just embrace it.

I'm not saying that it should be a complete free-for-all, rules could be put in place such as no overly explicit images if the images generated are photorealistic and no images of famous/well known/real people or copyrighted material.

The genie may not go back in the bottle, the case against copyright theft is yet to have its date in court. In the meantime generative AI is so hungry for images it is now scrapping AI images, this has the effect of poisoning the algorithm. Many graphics are being poisoned with Nightshade or Glaze software, more importantly the shareholders are getting tried of waiting for the wonder child AI to pay dividends.

 

[pixiemuledonkey]

I don't like AI, I'm going to say that first so hopefully I don't get as much hate. I also don't like how it is 'trained' though there is some misunderstanding about how it uses scraped data to 'create' new images.

I have prosopagnosia, face blindness. I am an artist but cannot create faces, nothing even partially realistic anyway. So when I needed a profile picture I used AI. My profile also clearly states that. I used ai once to create a basic concept for a shirt and gave that concept to an artist who did a stunning job making it perfect.

Anyone remember when digital art started becoming a thing in the early 90s? Do you also remember the fight between traditional and digital artists? We were convinced digital artists weren't real artists for a time but now we each have our place and no one is out of work because of it.

I'm not suggesting that AI is real art but I hope in a decade or less we can all settle into a comfortable place where AI art and real art live side by side. And that there are simple ways to determine if art is real or AI, just like there are ways to determine if a photo is shopped or not.

I'm interested to hear responses and I really hope it isn't attacks but I understand the anger, frustration and fear that is wrought by this new technology.

The introduction of digital art in the 90s was indeed met with resistance from some quarters of the traditional artists’ fields because it automated aspects of the work and required artists to purchase and learn to use the software and hardware in order to benefit. There was a perception that it would make it more difficult for wholly traditional artists to compete, or that it wasn’t true art because “the machine is doing all the work.”

Many of the arguments from that time do sound similar, I will admit. I think there are valid comparisons. For example, one can question whether the ability to automate aspects of the work reduces artists’ overall skillset because they may never learn to create those aspects on their own—I don’t know how to perform many photo techniques in an actual darkroom that I can do in Photoshop. There was also some legitimacy to asking what would ultimately be the cost to artists and art in general when tech corporations benefitted from the incorporation of their products into the creation of art as a perceived necessity. We’ve now arrived at a point decades later where digital art has made generative AI possible, allowing the tech industry to profit from the work of artists without compensation.

However, I dispute the notion that because digital art turned out to be “real art” after all that generative AI should be given the same benefit of the doubt. That conclusion strikes me as a relative of the proposition that generative AI does the same thing that human artists do when we look at the work of other artists and draw inspiration from them, adopting styles and techniques and incorporating them into their own work to create something new that is considered “original” despite being inspired by others.

Other people have detailed all of the flaws in this argument in far greater detail and more eloquently than I can—and have done so before U.S. Senate hearings—so instead of trying to reinvent the wheel here, what I want to challenge is the fundamental assumption underlying much of the pro-AI argument: that artificially generated stories and pictures are equal to human creativity and the works humans create. It’s a purely capitalist view of art as a commodity—“content”—for wealthy owners of AI to profit from without all of the messy consequences of dealing with human artists, such as having to compensate them for their labor or wait for said labors to be completed and made ready for sale. The entire purpose of genAI isn’t to make art more accessible to people, it’s to enrich and empower the already-rich owners of the technology. It doesn’t democratize art—art is already democratized, anyone can make it with any materials at hand—it further defines it as a product which can be bought and sold rather than an expression of humanity.

Furthermore, unlike digital art, the adoption of genAI by executives has already resulted in many artists being put out of work, worsening working conditions for artists because executives view “touching up” genAI results as a lesser service than providing original artwork and thus pay far less despite said “touch ups” often requiring an equal amount of labor. Executives thus feel justified and empowered to pay artists less for the same amount of work.

And we can all see how the ease and speed of getting genAI output has empowered scammers and made searching for erotica on sites like Amazon a hellish experience.

I think that conceptually there is a place for genAI as a tool for creatives. But in order for that to happen it has to be under conditions that don’t currently exist and strike me as unlikely to exist while genAI and American capitalism work the way they do now.

 

[blondebreeze]

The introduction of digital art in the 90s was indeed met with resistance from some quarters of the traditional artists’ fields because it automated aspects of the work and required artists to purchase and learn to use the software and hardware in order to benefit. There was a perception that it would make it more difficult for wholly traditional artists to compete, or that it wasn’t true art because “the machine is doing all the work.”

Many of the arguments from that time do sound similar, I will admit. I think there are valid comparisons. For example, one can question whether the ability to automate aspects of the work reduces artists’ overall skillset because they may never learn to create those aspects on their own—I don’t know how to perform many photo techniques in an actual darkroom that I can do in Photoshop. There was also some legitimacy to asking what would ultimately be the cost to artists and art in general when tech corporations benefitted from the incorporation of their products into the creation of art as a perceived necessity. We’ve now arrived at a point decades later where digital art has made generative AI possible, allowing the tech industry to profit from the work of artists without compensation.

However, I dispute the notion that because digital art turned out to be “real art” after all that generative AI should be given the same benefit of the doubt. That conclusion strikes me as a relative of the proposition that generative AI does the same thing that human artists do when we look at the work of other artists and draw inspiration from them, adopting styles and techniques and incorporating them into their own work to create something new that is considered “original” despite being inspired by others.

Other people have detailed all of the flaws in this argument in far greater detail and more eloquently than I can—and have done so before U.S. Senate hearings—so instead of trying to reinvent the wheel here, what I want to challenge is the fundamental assumption underlying much of the pro-AI argument: that artificially generated stories and pictures are equal to human creativity and the works humans create. It’s a purely capitalist view of art as a commodity—“content”—for wealthy owners of AI to profit from without all of the messy consequences of dealing with human artists, such as having to compensate them for their labor or wait for said labors to be completed and made ready for sale. The entire purpose of genAI isn’t to make art more accessible to people, it’s to enrich and empower the already-rich owners of the technology. It doesn’t democratize art—art is already democratized, anyone can make it with any materials at hand—it further defines it as a product which can be bought and sold rather than an expression of humanity.

Furthermore, unlike digital art, the adoption of genAI by executives has already resulted in many artists being put out of work, worsening working conditions for artists because executives view “touching up” genAI results as a lesser service than providing original artwork and thus pay far less despite said “touch ups” often requiring an equal amount of labor. Executives thus feel justified and empowered to pay artists less for the same amount of work.

And we can all see how the ease and speed of getting genAI output has empowered scammers and made searching for erotica on sites like Amazon a hellish experience.

I think that conceptually there is a place for genAI as a tool for creatives. But in order for that to happen it has to be under conditions that don’t currently exist and strike me as unlikely to exist while genAI and American capitalism work the way they do now.

thanks for putting in the time to explain so graciously the many delicate points that the issue of AI image generation touches.

 

[blondebreeze]

Making Art, with pencil , brushes , paints and pigments requires not only the time and willingness to learn and practice , the enormous challenge to find and define your visual style and , most of all, the courage to expose -for all the world to see- something that comes from the depth of you, that contains a part of you, and that you try - as imperfectly well as you can- to put on paper .
I tried AI and honestly I found it's just seamless photoshopped collages that BY DEFINITION cannot produce anything meaningful, because it's just a pixel-prediction algoritm.
If you have an idea for an image there are so many ways to try and produce it!
Doing so with AI , to me, it's the embodyment of the instant gratification mindset of people who knows the price of everything and the cost of nothing.
It's an easy way to avoid the frustration of learning and failing a thousand times.
And it just take away all the fun in creating something.
As for every new technology, we seem to be using AI , for the main part, in the most idiotic ways.
Ask yourself why, Who benefits from this cheap-flat-image machine?

But hey, to whomever this may concern, you're free to delude yourself of exercising any artistic endeavor by using AI, when you don't even know the first thing about composition. (And it's the exact same thing with the written word)

Artworks and stories are , in a way, precious gems that the artist goes to great trouble to mine from his soul , mind and hands.
Please don't compare it to sequins and rhinestones.

Just observe the children under 7: when they are drawing all their attention and energy is caught up in that moment.
What they produce is unfiltered , and it's them and their experiences and imagination. Would anyone advocate we should be teaching them to use AI?

Let's look at what science has to say about how our brain works:
Neurology already says that looking at a real painting, in a museum stimulates the brain far more than looking at a digital replica of it. By extension, I say, at any digital image. "Scientists in Netherlands using eye-tracking and MRI scans found ‘enormous difference’ between genuine works and posters"

As for the blanket site-policy, it's obviously the less painful road at the moment, it keeps them out of so much trouble.

And , for all those who give a fxxx about this earth, AI usage TODAY is already generating as much pollution as the Netherlands.
"The proliferating data centres that house AI servers produce electronic waste. They are large consumers of water, which is becoming scarce in many places. They rely on critical minerals and rare elements, which are often mined unsustainably.
from UN website

 

[000zing]

Seconded.

We live in a cheapened, tawdry age where human endeavour is derided and the machine is God. Allowing this to happen will come at a high price... quite soon... and to us all. Enjoy "your" creations while you can.

 

[Melody_Lightdraft2]

What we need is a conclusion of the copyright case brought by creative artists against the Generative AI companies.
I mainly use Daz3d software for my picture prompts and graphics.
With AI, it's the hands that give them away, or tack on equines.

 

[Sappho_CGN]

I think, the site has also legal reasons for avoiding AI use.

As I understand the process, the AI takes elemetns out of texts or pictures in the internet and combines that to a new text or picture. So you can have copyright problems because parts are "stolen" from other art etc.

Finally, I tested text AI .... to be honest, the outcome is horrible. Maybe such AI can set up casual textes but thats it.

 

[Otto26]

Quantity over quality - The example of DeviantArt was raised. Genuine art stands out from the crowd of AI tripe. I think it always will. This goes for written works as well as visual works.

Copyright - The only good argument I have seen. I have yet to find a generative AI model that didn't incorporate copyrighted works without the permission of the creator. On the other hand, you have the example of Content Aggregation sites on the internet which pull together images and text from other sources, frequently without permission or proper attribution. And this site which says you have to own the copyright hosts thousands of pictures to which the posters obviously do not own the copyright and publishes 2D images of 3D renders which were obviously produced in online gaming platforms or using render tools. I understand that the site owners wish to encourage individual creativity and avoid copyright issues and are very, very busy but I wish they'd either apply the policy equally or modify it to reflect reality.

Pay an artist - Not for free porn being published on the internet. Not to mention that the few times I have paid an artist they have never produced what I had in my head, not matter how many descriptions or sketches I provided. If I'm running a for-profit business then yes, absolutely, pay an artist.

If it's not handmade it's not art - I think this is a load of ****. I've used electronic tools to produce original and reproduction art for years and it has been stomped on by amateurs because they can't stomp on the companies that are normalizing this practice. Digital art has been brought up and artists were opposed to that. Which is normal. Artists opposed the adoption of the printing press.

Inevitability - Remember how I said artists didn't produce what I had in my head. AI can. Businesses with deep pockets are going to monetize the heck out of generative AI and the absolute best that artists can hope for is that they are forced to do so with open source artwork or artwork for which they have paid for the rights. How many of you are going to be compensated in that dream scenario? The same is going to hold true for LLMs that generate text. When the day comes that you can generate a decent porn story on a readily accessible LLM at least 50% of the market is going to have the tool generate what they specifically want to read. But, as I said earlier, human generated content will always stand out.

Warhol - Someone brought this up and it's a great example. Consider the Campbell's Soup portraits. Warhol used a projector to cast the images on canvas and traced over those and then hand-painted them. Should we not consider that art? And there's a long, long tradition of tracing existing art and repurposing it in new pieces of art that goes back at least as far as the manuscripts of the middle ages. Warhol's point was that art shouldn't be the exclusive domain of a select few, both the consumer and the creator, and that art didn't have to be some great, original work. He also didn't have the permission of whatever artists came up with the soup label designs. I've made composite/collage images using images in the public domain and had them rejected by this site because I didn't create them when, in fact, I did create new content using images I had the right to use.

Fight the change - The buggy-whip industry would like to speak with you. Adapt or die. Seriously. Generative AI isn't going anywhere and the trick is to figure out how to live with it, not how to make everyone give it up.

 

[gsnowpainter]

It is not art in that it is not creation of images using one's own hand and talent. It is entering prompts and letting a machine do the work.

I see the problem as a too broad application of AI and art. Andy Warhol used photos of famous people and, wwith and without their permission. The Campbell Soup Company never complained. All he did was make serigraphs of the orginal image and print those. Ant curator and art educator will tell you that it is art! Besides, we have seen very artistic Ai produced art and also some very poor ones. THere must be some skills invol
ved.

 

[gsnowpainter]

I think, the site has also legal reasons for avoiding AI use.

As I understand the process, the AI takes elemetns out of texts or pictures in the internet and combines that to a new text or picture. So you can have copyright problems because parts are "stolen" from other art etc.

Finally, I tested text AI .... to be honest, the outcome is horrible. Maybe such AI can set up casual textes but thats it.

I think that you are spot on re legal consideration: a fear of possibly being sued. My main objection is the " one size fits all" appication.

 

[DiscipleN]

I've got no horse in this race. I do find the hysteria - familiar.

from: https://www.nga.gov/features/in-light-of-the-past/the-19th-century-the-invention-of-photography.html

In 1839 a new means of visual representation was announced to a startled world: photography. Although the medium was immediately and enthusiastically embraced by the public at large, photographers themselves spent the ensuing decades experimenting with techniques and debating the nature of this new invention. The works in this section suggest the range of questions addressed by these earliest practitioners. Was photography best understood as an art or a science? What subjects should photographs depict, what purpose should they serve, and what should they look like? Should photographers work within the aesthetics established in other arts, such as painting, or explore characteristics that seemed unique to the medium? This first generation of photographers became part scientists as they mastered a baffling array of new processes and learned how to handle their equipment and material. Yet they also grappled with aesthetic issues, such as how to convey the tone, texture, and detail of multicolored reality in a monochrome medium. They often explored the same subjects that had fascinated artists for centuries — portraits, landscapes, genre scenes, and still lifes — but they also discovered and exploited the distinctive ways in which the camera frames and presents the world.

 

[gsnowpainter]

I've got no horse in this race. I do find the hysteria - familiar.

from: https://www.nga.gov/features/in-light-of-the-past/the-19th-century-the-invention-of-photography.html

I agree in every way, as did Claude Monet and others.

 

[OffNSFW]

I, too, find it familiar, almost as if the topic has already been discussed to exhaustion in about 10 other threads.

Learn to draw. It's good for you.