That photo is horrendous, so I sure hope for the artists sake it was a 30s prompt job.
I will push back slightly on the idea that slop is only slop because it's bad. AI art will always be slop because of its very nature, it's not even possible to describe ai art as "good" because that's not a quality it can even possess.
"It either is AI-generated slop or it looks like AI-generated slop for no artistic or thematic reason whatsoever."
The artistic and thematic reason is obvious. It's a commentary on the show and AI art in general. To ignore this message because it "looks like AI" devalues the entire concept of human art.
Everyone in the industry who is trying to find a way to weasel and deny using AI is using the same "incorporate standard digital tools" bullshit phrase. I think Epic said the exact same phrase when the Fortnite community caught them using AI art in the latest season of the game.
The "non-denial denial" seems like a denial to me if you take him at his word that he is unable (presumably due to legal reasons) to comment about a specific project.
I have a very sensitive ai-radar and I don't think this illustration is AI. It has an extremely uncanny look to it, but totally unlike the uncanny sameness of AI slop. I think it's highly unlikely it's an AI image, especially coupled with the denial by the artist.
False AI accusations are very harmful, especially when so much actual AI slop goes unnoticed.
It's labeled "whole milk" in small text then below in bigger text it's labeled "low fat milk". The fact that this was not caught and likely would not be deliberate by the author seems slopy (sloppy?) to me. Just my two cents
Isnt it an intentional choice? The show Pluribus has a not-so-subtle analogy to an LLM that has access to the world’s knowledge, doing its best to give whatever we ask it for, and not being the brightest tool in the shed. I think the artist accomplished what they wanted to in the context of the show.
EDIT: I stand corrected, the show’s analogy to AI is coincidental.
2) Two of the maze walls appear to be smudged, but nothing else in the image is.
3) The floral pattern on the plate is messed up - it repeats, but berries randomly change positions and the leaves change shape.
4) The milk carton has a splotchy texture that's really common in ChatGPT-generated illustrations.
I suspect it's not 100% gen AI - for example, the sharp outline around an out-of-focus xmas tree feels like a human-made artistic choice - but I'd bet it's at least partly composed from AI-generated elements.
Surely the fact that "It says low-fat milk and whole milk; no one would do that" is followed by "oh the real props do that" should cause everyone who thought the former to down-rank their own ability to tell what's AI and what's not. If they don't then they aren't incorporating evidence about the world and their skill.
To have a separate conversation from everyone else (who is talking about whether it's real or AI), I think it's interesting to see people's epistemology. If you thought something was X because of A (i.e. P(X|A) > P(X), maybe much greater), your posterior for P(X|A) should be different from your prior in response to the evidence "it was X, but not X also has A" and I think the directionality should be obvious.
For those who don't do that, I should update my adjustment factor to their claims of fact and not in their favour.
These artists who do work for tech companies are literally given a day to do this work. Of course they are using AI. They have to. The days of having time to make something great are mostly over.
20 comments
[ 3.6 ms ] story [ 42.4 ms ] threadI will push back slightly on the idea that slop is only slop because it's bad. AI art will always be slop because of its very nature, it's not even possible to describe ai art as "good" because that's not a quality it can even possess.
The artistic and thematic reason is obvious. It's a commentary on the show and AI art in general. To ignore this message because it "looks like AI" devalues the entire concept of human art.
He’s reaching to try and find any leeway to say it was AI based on the words the artist used.
But then he ends with “well it sucks anyway”
Dude, just say “my bad” and “it’s not to my tastes”. You don’t need to double down so hard.
False AI accusations are very harmful, especially when so much actual AI slop goes unnoticed.
EDIT: I stand corrected, the show’s analogy to AI is coincidental.
1) The maze on the carton can't be solved.
2) Two of the maze walls appear to be smudged, but nothing else in the image is.
3) The floral pattern on the plate is messed up - it repeats, but berries randomly change positions and the leaves change shape.
4) The milk carton has a splotchy texture that's really common in ChatGPT-generated illustrations.
I suspect it's not 100% gen AI - for example, the sharp outline around an out-of-focus xmas tree feels like a human-made artistic choice - but I'd bet it's at least partly composed from AI-generated elements.
And for what it's worth, the image pretty consistently trips gen AI image detectors (e.g., https://hivemoderation.com/ai-generated-content-detection).
To have a separate conversation from everyone else (who is talking about whether it's real or AI), I think it's interesting to see people's epistemology. If you thought something was X because of A (i.e. P(X|A) > P(X), maybe much greater), your posterior for P(X|A) should be different from your prior in response to the evidence "it was X, but not X also has A" and I think the directionality should be obvious.
For those who don't do that, I should update my adjustment factor to their claims of fact and not in their favour.
Shark is just a shark. So smart, me!
:)