they are trying to claim they made something but instead what they really did was go to chat GPT or codex or something like that, and spend 5 minutes getting it to make a bunch of images and claim that this is worth $5 or $15.
While I agree that the content should be appropriately tagged, your claim also doesn't show reality as is, and won't be a surprise if many of those who upload AI assisted artwork decides not to apply the tag out of some kind of protest.
Sure, I don't know how they made their content , I can only talk for myself. And I will do that. Let me ask you first, how you think that works? You just say chatGPT 'hey made me a whole game' and it spit out a ready to play .exe file? Or 'hey computah... make me assets' and it spit out a bunch of .blend files?
It doesn't work like that. You may see in YT or other socials people showing they do stuff mostly with AI but those games are just more like basic prototypes of 'proof of concepts' rather than a fully playable and ready to ship games. And If the walk about the assets, same thing.
I used AI in some of the assets i upload here. Sometimes for up-scaling (taking a image of x resolution and produce the same in higher one without losing quality) sometimes for refinement (drawing the texture by hand (or well... traditional methods) and using AI, local AI BTW, to polish it to produce a more photorealistic albedo map, that i have to draw and re-refine over and over until reaching the desired result).
I'm recently using huanyang-3d (I think i wrote it wrong but anyway) to locally generate meshes from images. It's... okay, more or less, the meshes required polish and texturing work (it doesn't generate textured meshes) but I kinda developed a workflow that allows me to create a fully textured character in one afternoon more or less. How? by combining several separated steps. Creating a character image with limbs spread, removing the background, separating in pieces (head, hands, body, etc.), using the hunyuan-3d to generate a mesh from each image,then in blender do a remesh to lower the triangles count to something actually usable for each part, mark seams, unwrapping and all, then, also using another tool i generate a basic texture over the topology heavy one and bake it to the lower one, then in Krita with the plugin 'Krita AI Diffusion' I fix the texture, redraw parts, edit the image overall to get the finished albedo map (and this is for just one of the body parts) then i have to make sure the texture fits right in the low triangles model -> re-modify what's needed (either mesh or image)-> re-verify, until it's good. Merge all body parts into one, sometimes using a topology bridge script I have (made with chatGPT). And each part, requires lot of work to do by hand, directly interacting with the PC, you know, 'effort'.
So yeah, If many other devs are facing this same scenario, doing a lot of work to then be ignored or cursed upon or whatever, then it's fully understandable if they don't tag their work as AI assisted. And I didn't mention something more ironic, imagine doing a fully textured 3d model by hand, textures from baked shaders or photos, with normals and etc, but then having to tag the work as AI-assisted only because you used a small AI texture in a corner because you couldn't find a fitting texture anywhere and boing by hand was time-consuming or really hard.
It's kinda understandable if some decides to ignore or not using the tag because in the end it doesn't feel fair. Imagine doing a lot of work only to be ignored or attached based in a process the other person doesn't fully understand, nor understand at all. (And let's face it, some antis are bloody delusional at this point.)
I still believe everybody should appropriately tag their work. Slop is slop no matter the tool used but if you think that the issue with the slop or the showelware is because AI. Well... Is like blaming the moon for the night's darkness.
