Modern "art"work redone by gpt-image-2 but they pretend that a human did that “manually” in Photoshop and illustrator
Origin: https://system3.com/news/from-retro-pixels-to-the-pondversethe-bovine-bomber-recreated-in-3d/
“Human-led” process, it was called in some other article.
Do you by any chance have a link to said article?
Human led process then don’t blatantly lie in the marketing copy where they pretend to redone everything manually in illustrator
Do you think if i started rewriting books from the 1990s, but updating the slang and such, would people give me money to buy food? Ha
redone by gpt-image-2
Source? The article you linked says:
No generative AI was used in the creation of the artwork. Every stage began with the authentic original James Pond graphic and was completed using traditional digital illustration, 3D rendering and manual artistic refinement.
So if you have proof, please link the article which says it’s genAI, not the marketing one.
No proof but the style is reminiscent. On top of the fact that one of the horns is missing, the arc disappeared from its head, one of the udder rockets has been dislodged, and a pivot disappeared in the leg.
The horn is only missing in the 3D render, not the final photoshopped result. Actually this might be a case of “ugh let’s just add this in post” instead of them having to model the horn. The arc disappeares because it’s a 3D render and it’s annoying to create lightning in 3D, much easier in 2D. The rocket looks like it’s perspective and I don’t see what you mean with the pivot.
Year that’s the point. The “3D render” is the genAI model.
I don’t believe it, but you can prove me wrong by using the chatgpt-2 thingy and generating it again. Bet it looks totally different.
It will look different. It always does.
That’s the style of gpt-image-2. Without doubt. Source: all the thousands of pieces of slop with this style
Replace shopped with slopped and there’s a clean meme
Bruh. Of course it looks like AI because AI is mimicking 3D artist’s works.
I’m really against AI “art” as it’s unimaginative, derivative, morally bankrupt and frankly offensive. But we should be wary of calling real art “AI” just because we don’t like it, as this is a serious insult.
I once read of a non-AI artist being bullied on Twitter to the point they deleted their account and harmed themselves. This is something I will never want to be a part of. In dubio pro reo, especially since the studio in question seems to be quite proud they didn’t make it with AI.
I’ll be careful here and say it’s not AI, even though I don’t like the style either.
It’s the second image, not the 3d render. It’s impossible that in 2026 a human artist will intentionally draw in that style letting everyone think that it’s ai generated.
All cartoon characters generated by gpt are in that style
So an artist’s style becomes forfeit because AI has copied it?
This also assumes that an artist will just abandon their style because AI was trained on it. Artists’ styles are developed and grown, really part of the artist themselves - it can’t just be swapped out like a module.
All up it’s a pretty shit take, you night want to think through it more thoroughly.
in this specific case, the image is genai, period. I saw enough of the same stuff to distinguish it. It’s the same photocopied style all over. If system3 wanted to show their “handmade” process, they should have shown a pencil draft or so, like they did until 3 years ago. also: what kind of process, someone does it in photoshop (raster) AND illustrator (vector) - it’s supposed to be only a reference for the 3d render, but they added the weird shading.
the series creator said it’s genai too: https://www.engadget.com/2245979/james-pond-creator-is-not-happy-with-the-series-allegedly-ai-generated-legacy-collection/
Aha, thank you! That’s a way better source than the initial article.
Somehow, the 3D version looks more like the original than the 2D version, but it still has that fourth teat that’s not attached to anything…
