As AI-generated text continues to evolve, distinguishing it from human-authored content has become increasingly difficult. This study examined whether non-expert readers could reliably differentiate between AI-generated poems and those written by well-known human poets. We conducted two experiments with non-expert poetry readers and found that participants performed below chance levels in identifying AI-generated poems (46.6% accuracy, χ2(1, N = 16,340) = 75.13, p < 0.0001). Notably, participants were more likely to judge AI-generated poems as human-authored than actual human-authored poems (χ2(2, N = 16,340) = 247.04, p < 0.0001). We found that AI-generated poems were rated more favorably in qualities such as rhythm and beauty, and that this contributed to their mistaken identification as human-authored. Our findings suggest that participants employed shared yet flawed heuristics to differentiate AI from human poetry: the simplicity of AI-generated poems may be easier for non-experts to understand, leading them to prefer AI-generated poetry and misinterpret the complexity of human poems as incoherence generated by AI.
The thing I really hate about AI is when they say it can make art. For centuries, art has been a form of expression and communicating all sorts of human emotions and experiences. Some art reflects pain or memories experienced in life. Other art is designed out of intellectual curiosity or to evoke thought. AI isn’t human, so it can’t do anything other than copy or simulate. It’s artificial after all. So it makes images. But there’s no backstory or feelings or emotion or suffering. It’s truly meaningless.
In 1962 Phillip K Dick put out a book called “Man in the High Castle.” In it there was a scene that stuck out to me, and seems more and more relevant as this AI wave continues.
In it a man has two identical lighters. Each made in the same year by the same manufacturer. But one was priceless and one was worthless.
The priceless one was owned by Abraham Lincoln and was in his pocket on the night he was assassinated. He had a letter of certification as such, and could trace the ownership all the way back to that night.
And he takes them both and mixes them up and asks which is the one with value. If you can no longer discern the one with “historicity,” then where is it’s value?
And every time I see an article like this I can’t help but think about that. If I tell you about the life and hardship of an artist, and then present you two poems, one that he wrote and one that was spit out by an LLM, and you cannot determine which has the true hardship and emotion tied to it, then which has value? What if I killed the artist before he could reveal which one was the “true” poem? How do you know which is a powerful expression of the artist’s oppression, and which is worthless, randomly generated swill?
Art, like the value of Lincoln’s lighter, is in the eye of the beholder.
Often, people find art in completely natural occurrences. Or even human designs seen in certain ways, like how two or more separate buildings might come together in unintended ways.
So, even if it’s not strictly intentional human art, it’s still valid to appreciate it.
There’s no contradiction here.
With high value art you definitionally buy a story not the content. Without a certificate of authenticity or a story that goes with it there is no story and no value to it.
With K Dick’s example the two lighters would become of different but equivalent value, perhaps the new value is in the story of how two identical copies and yet different came to be.
You could 3d scan the statue of David and reproduce it down to its tiniest detail. And yet the copy is only worth as much as the cost to make it or even less, while the original is invaluable.
You can see the Mona Lisa on your phone any time you want and yet millions will take the trip to the Louvre to see what is most likely not even the original.
The story and the history of an object is what you purchase when buying art or antiques of high value.
They didn’t say AI produces low value art. They said AI doesn’t produce art at all.
I think there’s an argument about art being the emotions it invokes in the viewer rather than the creator. Humans can find art in natural phenomena, which also has no feelings or backstory involved.
I’m not really defending AI slop here, just disagreeing with your definition of art and the relation to the creator rather than the viewer.
Indeed, there are whole categories of art such as “found art” or the abstract stuff that involves throwing splats of paint at things that can’t really convey the intent of the artist because the artist wasn’t involved in specifying how it looked in the first place. The artist is more like the “first viewer” of those particular art pieces, they do or find a thing and then decide “that means something” after the fact.
It’s entirely possible to do that with something AI generated. Algorithmic art goes way back. Lots of people find graphs of the Mandelbrot Set to be beautiful.
Or, maybe, we have to accept that art and all the grandiose and deep narratives around it are bullshit. It’s an illusion, it’s just a tool so some of us feel more important.
All that crap about not being made by humans is just the fear that the illusion of grandeur of humans might collapse.
I do get the sense sometimes that the more extreme anti-AI screeds I’ve come across have the feel of narcissistic rage about them. The recognition of AI art threatens things that we’ve told ourselves are “special” about us.
Correct.
And especially artists, or people aspiring to be artistic, are suffering from an inferiority complex which they try to hide behind grandiose “higher values” of art.
AI threatens to expose that art is meaningless unless you can use it to distinguish yourself from the plebs, or those you deem plebs.
Art is in the act of creating it. Not in the final product to be bought and sold on the market.
A kid coloring is making art. The joy they get in the making is the art and is the point.
I feel sorry for so many people in this thread who keep approaching this from the point of view of consumer markets. It doesn’t matter if someone can determine an AI colored picture from a child’s. The AI derives no joy in the creation. It’s not art, but a copy.
This is just your opinion.
This isn’t inevitable or necessary.
Personally I enjoy generated art (mostly scifi/fantasy) and I never pay for it.
On the other hand I try to support actual artists because they’re most often struggling under capitalism much more than some person using midjourney or wahtever.
There’s a lot of consumer/commodity notions about art in this thread.
I write poetry because self-expression helps me appreciate life more deeply. I share my self-expression with others who will appreciate it. Mostly, people who know me personally and other poets.
Art is soul food. Until machines realize they exist, and one day will not exist, they can’t self-express, and aren’t doing art.
They can imitate it well enough to fool consumers. But that doesn’t make it art.
To quote one of my favorite lines, sticking feathers up your ass does not make you a chicken.
I think Lemmy’s general demographic skews towards techy early-adopters and lots of STEM background folks and it shows with topics like this. I’m not saying that’s a negative thing, just that it’s the vibe here.
Art is just such a broad topic, it gets messy. Plus I think the verbage around discussing it isn’t as universally defined as in other topics. It doesn’t always fit neatly into categories and boxes that can make it harder to have nuanced discussions.
There’s no such thing as “AI”.
But computers can also generate art through averaging. It can average the feelings, impact, etc. That’s part of why generated art is popular. It’s still people creating new works from the old. It’s still “art” by any reasonable definition.
How do you mean there’s no such thing as AI?