AI/LLMs AI art bans are going to ruin small 3rd party creators


log in or register to remove this ad


Yes, and posts have context. The context of my post was first and foremost comparing LLM 'art' with similar people created art. As such only the digital definition needs to be applied.
So? It was a poor definition that does not match how the word is used. The context is irrelevant, you're used a word with a definition and then attempted to re-define it in such a way to exclude things that people unequivocally call art, all to try to prove a point about AI.

I can describe food as something between two slices of bread, but that not only isn't right due to the overwhelming evidence of other food items, it also doesn't mean if I stick a brick between two slices of bread, its suddenly food. You can't just redefine common terms that hard.

For the kinds of art my definition is meant to apply to, it is very useful and absolutely correct. Insisting that one must always use a more universal definition when talking specific subsets in a specific context is not logical.
You didn't say 'kinds of art'. You said 'art' in its entirety. And its not useful, nor is it correct. Me slapping keys on a keyboard is creating pixels on the screen, but I wouldn't say the majority of my forum posting is art, even if I occasionally come up with some bangers in Discord.

Likewise myself going into Minecraft and painstakingly using my gathered supplies to create a castle easily could be said to be art, but I'm not creating any pixels at all in that process. If you're arguing that Minecraft builds aren't art, I've got questions

And heck, I haven't even gotten into music. "Creating pixels" would mean that music isn't art by this definition.

Its a terrible definition that doesn't fit any description of art, that you're trying to hammer down to prove a point about AI.
 

The context is irrelevant,
The context is always relevant.
you're used a word with a definition and then attempted to re-define it in such a way to exclude things that people unequivocally call art, all to try to prove a point about AI.
No I didn't. I defined what digital art is, the only thing LLM's can actually produce, which is the key context of this whole discussion.

Like why the heck does it matter if art is also painting a physical paint brush on a physical canvas? Why does that have anything to do with the topic of whether a human can use a LLM as a tool to output art (oh wait, not art, digital art, I wouldn't want to be accused of redefining a word)
I can describe food as something between two slices of bread, but that not only isn't right due to the overwhelming evidence of other food items, it also doesn't mean if I stick a brick between two slices of bread, its suddenly food. You can't just redefine common terms that hard.
My definition wasn't incorrect for digital art though.

You didn't say 'kinds of art'. You said 'art' in its entirety.
Okay. Can you accept that I thought the context should have made the implication and limitations apparent? If so, can you see why I think it fine to use the broader term art in place of the digital art subset given the context?

Likewise myself going into Minecraft and painstakingly using my gathered supplies to create a castle easily could be said to be art, but I'm not creating any pixels at all in that process. If you're arguing that Minecraft builds aren't art, I've got questions
Wasn't my definition creative intent alongside pixels being produced makes art? It's hard for me to understand why you think that wouldn't apply to minecraft? But in my opinion that would be art, sorry digitial art, and that clearly falls under the definition.

By the way, why did you call it art here instead of digital art? Are you redefining the word now? ;)
And heck, I haven't even gotten into music. "Creating pixels" would mean that music isn't art by this definition.

Its a terrible definition that doesn't fit any description of art, that you're trying to hammer down to prove a point about AI.
Okay music isn't included in this definition. You got me ;) But, so what? (nevermind that this whole discussion context has thus far been about LLM digital images).

I don't understand why instead of discussing my point where it does apply, that you are trying to point out places where it doesn't? What's the purpose there?

Like assume for a moment that my definition isn't broad enough and the context isn't clear enough. Does that really matter so long as it is accurate for some subset of art? The meat and potatoes is whether it's accurate for some subset of art not whether it should be broader or is accurate of all subsets of art. That is, unless it's not accurate for every subset of art, then it would still show how LLM's can be used by a human to produce art in at least one subset of art.
 
Last edited:

I think art is when pixels get created in response to human creative intent.
A prompt engineer doesn't create pixels any more than a recipe writer creates food. A prompt engineer creates prompts. Any pixels an AI creates based on a prompt one feeds into it are created by software with no human creative intent, not by the prompt writer.

Likewise, if someone in the future invents an automated kitchen, a recipe writer still doesn't produce food. A recipe writer produces recipes. Any food an automated kitchen produces based on a recipe one feeds into it is created by robots with no human creative intent, not by the recipe writer.

(The above remains true even if the recipe writer tastes the food produced by the automated kitchen, writes an improved recipe to adjust the flavor, and iterates the process in a back-and-forth conversation between the written recipes and the automated kitchen. Despite all the human creativity the writer puts into the iterated recipes, none of the writer's actions are producing any food. The kitchen is producing the food.)
 

What if Andy Warhol used an AI art prompt to describe and tweak his vision for how he could alter soup cans to make a statement and the output ended up exactly the same which he then transferred to canvas via a printer? Would that then make it not art?

What if Andy Warhol had been dead for almost 40 years?
What if Andy Warhol was the only genius to have had a 60 IQ?
What if we talked about reality instead of hypotheticals?

In what way is the vision being tweaked in your example? And for what reason? To say what? These are the important parts. The context.

Readymade art, both in Dada, (the movement that spawned it) and by the pop artists that revived it, were interested in seeing the beauty in everyday objects that people take for granted. Giving them a new context by placing them in a gallery. In most art movements it's the thinking behind the work that's probably more important than the work itself. Which is the issue with AI as an art movement so far, it's all image and no thinking.
 
Last edited:

A prompt engineer doesn't create pixels any more than a recipe writer creates food.
Assuming the prompt engineer is hitting run on the prompt then he is doing something fundamentally different than the recipe writer who simply writes a recipe and does nothing more.
A prompt engineer creates prompts. Any pixels an AI creates based on a prompt one feeds into it are created by software with no human creative intent, not by the prompt writer.
You seem to be assuming, AI produces single image, prompt engineer accepts that image. However, the more common flow is, AI produces image, prompt engineer asks for modified image as it's still not quite to his vision and it becomes an iterative process.

The creative intent need not be in the software alogrithm itself, but it is clearly in that iterative process that was used to create the final image.

Likewise, if someone in the future invents an automated kitchen, a recipe writer still doesn't produce food. A recipe writer produces recipes.
Sure, but again, the prompt engineer doesn't just give a recipe but also says run. That's fundamentally different than a recipe writer which never tells anything/anyone to actually produce his recipe.

Any food an automated kitchen produces based on a recipe one feeds into it is created by robots with no human creative intent, not by the recipe writer.
Assuming you mean a person provides the automated kitchen a recipe and says make this, then it's a bold claim to say the philosophy of causality would not assign ultimate responsibility for that meal on the recipe writer who kicked off the mechanical process of production.
(The above remains true even if the recipe writer tastes the food produced by the automated kitchen, writes an improved recipe to adjust the flavor, and iterates the process in a back-and-forth conversation between the written recipes and the automated kitchen. Despite all the human creativity the writer puts into the iterated recipes, none of the writer's actions are producing any food. The kitchen is producing the food.)
I think we are at a fundamental disagreement on this, the details better explained earlier in this post.
 

Stock art, stock art, get stock art. I keep seeing this.

The question hasn't been raised: what if I, as a creator don't want or like stock art?

You're basically saying I should get stock art even if I don't like it because it's better than AI, and if I do get AI because I prefer it to stock art, I'm a bad creative at best, a bad person at worst.

Do y'all read this sh**t back to yourselves sometimes?
 

Stock art, stock art, get stock art. I keep seeing this.

The question hasn't been raised: what if I, as a creator don't want or like stock art?

You're basically saying I should get stock art even if I don't like it because it's better than AI, and if I do get AI because I prefer it to stock art, I'm a bad creative at best, a bad person at worst.

Do y'all read this sh**t back to yourselves sometimes?
If you want something you like more than stock art, hire an artist to create it. You can even give them feedback to iterate on the art so it's even more to your liking. Or you could create the art yourself!

Or you could use AI and choose to not sell your work on Foundry or other sites that don't allow AI products to be sold.

You could also choose to just steal art from an artist and not give them credit. That's a bad choice, but so is using AI.

You have a lot of choices!
 

Stock art, stock art, get stock art. I keep seeing this.

The question hasn't been raised: what if I, as a creator don't want or like stock art?

You're basically saying I should get stock art even if I don't like it because it's better than AI, and if I do get AI because I prefer it to stock art, I'm a bad creative at best, a bad person at worst.

Do y'all read this sh**t back to yourselves sometimes?
First post so I'm sure you're on the level but here are some options you missed:

  1. No art
  2. Art drawn by you
  3. Art drawn by a friend or playtester
  4. Art that isn't art (e.g. photos)
  5. Art funded by you
  6. Art funded by a kickstarter
  7. Art funded by making something smaller first
  8. Don't publish
 

Recent & Upcoming Releases

Remove ads

Top