On the contrary; it seems to me that caring about its authorship even though the product may be outwardly indistinguishable from something a human could have made shows that human authorship is, in fact, precisely what they value.
Are you trying to make an argument that "if they really valued human authorship so highly, they'd devote their every waking moment to learning what every possible method is to tell when a given work was created by an LLM" or something? Because I would think that was fairly obviously bullshit by this point. For one thing, there's plenty of ways to make LLM-generated works (especially smaller, simpler ones) that are 100% indistinguishable from something human-created. (This is especially true in text formats, where the amount of information, and thus the possible failure modes, available in the output is smaller than in images or video.)
I don’t want something you prompted an AI for because genuinely I could just do that myself. The barrier to entry is nil. I’m not interested in the idea an artist has I’m interested in how they execute it.
If they executed it through human effort then I place a much much higher value on that. Same as I would if someone selling a hand forged a knife rather than a mass produced cut out.
Even if the human-made item is lower quality I’d prefer it by a mile if for no other reason than I know someone’s love and effort went into it.
I could never have that appreciation for something a machine generated.
The lines might blur if an artist uses an AI to manipulate their own art that they made.
Point is - the output quality fundamentally doesn’t matter. It’s principle, not the output quality, that guides most people in avoiding AI art.
> The lines might blur if an artist uses an AI to manipulate their own art that they made.
Where’s the line? Is using tools in maya and photoshop to rapidly speed up asset generation ok?
> I could never have that appreciation for something a machine generated.
Do you feel the same about PCG?
By that logic, the fact that (some) people can't tell between gold and fool's gold is strong evidence that gold was not what they value at all.
Or: the fact that (some) people can't tell between food that is marketed to them as "healthy" and food that actually makes them healthy, is strong evidence that healthy food was not what they value at all.
Neither follows. Both blame the victim of a lie.
Well, no? Most people don't value gold for intrinsic attributes of the metal, but for the market-assigned value of it or perceived status.
I think GP overstated the point, but it's clear that a lot of people who object to AI authored content are objecting not because of the intrinsic attributes of the content but because of their emotions on the subject.
That doesn't make it invalid. If someone really doesn't want to consume AI generated content, then so be it.
But I also think the people making a big deal out of this seriously overestimate how deeply most people care. I suspect most people care only to the extent that the quality is lacking.
I'd justify that view with your example: People keep eating unhealthy food despite knowing it's bad. And as another example, we keep buying things we can't be sure is ethically sourced. We're willing to look past a whole lot.
The same way "Made in ..." labels were originally protectionist measures, I suspect disclosure requirements etc. will seriously backfire when it turns out most people stop caring the moment the quality is usually good enough.
See e.g. corporate green and rainbow washing, Amazon marketplace, organic food.
Do you want to have to go ask for refund after refund after noticing it's yet-another-AI-slop-cookie-cutter-lookalike several hours in?
If you buy a fake that looks extremely close but falls apart a month later, that doesn't mean you don't care.
If art is visually polished in some way such that quick quality cues are handled that doesn't mean that it will actually have a meaning or a message. You won't know until you have a chance to fully consume it.
It's a Turing test for art -- and the machines are winning, at scale.
And code indeed has its values. See all algorithms invented by Carmack like BSPs, z-fail, etc.
The problem is very few (none?) of them actually were. He was just a wizard at applying them. This furthers the utilitarian argument.
I think he did come up with "Carmack's Reverse" (a dynamic lighting algorithm) independently. But not even the "fast inverse square root" was his; that, oddly enough, came from what is currently probably the premiere AI company: NVIDIA.
But his implementations of these algorithms were indeed beautiful. I suppose AI could do something similar today, because the likes of Carmack had come before. But using AI for code is hobbling the creation of new Carmacks; and humans have the remarkable property that they can learn from other humans and still push the boundaries forward. When AIs learn from other AIs, model collapse occurs.
I don't think is so. Models continuing to improve by learning from other models is a counterexample.
The art and audio is somehow seen as more important and something only humans should de in a game than the code and design, you know the actual game portion of the game.
Neither of these changes affect the game play, they are both purely aesthetic in their effects, without introducing anything which might be called "art" in a traditional sense, unless you count crudely modeling the behavior of light in shaders and C code as "art", but at that point the distinction between code and art vanishes. I had previously had a "no-ai" policy in my codebase, just as a kind of over-reaction, and kind of, "if I ban it, then I don't have to think about it" response. But these patches... after studying them, there's just no way I wasn't accepting them.
[1] https://github.com/smcameron/space-nerds-in-space [2] https://en.wikipedia.org/wiki/Shadow_mapping [3] https://en.wikipedia.org/wiki/Gravitational_lens
You can't force people to like it.
And those same people don’t care about the AI code in the same games.
It's not and never will be.
There's no good reason for it.
Nobody seems to piss on an artist who leans on AI for the programming side of their game, whereas when it's the other way around...
The consistent position is to condemn all of it, or to only focus on the quality of the outputs. AI-gen art looks like shit, and always requires an artist to fix it post. (Just like letting an AI vibe-code it's way to a finished program with no programmer oversight will inevitably produce a large pile of shit.)
At least in certain genres, the use of AI art (or worse, writing) can become obvious to the player, and it winds up breaking whatever level of immersion that existed.
There's also the issue where it causes the 'signal/noise' ratio to go out of whack; Using Visual Novels as an example, the amount of shovelware has increased by at least one order of magnitude.
You can still have a VN with good enough writing, that 'AI Art' can be forgiven; and some devs actually do that to get started, i.e. they are hoping the project gets enough interest and support that they can go ahead and pay for an actual artist.
AI art is human art. It is based on human intent and choices (and in non-trivial cases it’s a lot of effort to get the machine to produce something that resembles a artist’s mental image), and the rest is mathematically sourced from human art.
I strongly suspect the only difference is in level of decision making, where details that aren’t explicitly planned are randomly generated rather than chosen (consciously or not). Or unable to be planned - refining the generation results is the most difficult part. Which translates to that “sloppy” image.
Humans can detect sloppiness, not “AI”.
There's an insane amount of fine-tunes out there (be it base models or LoRAs), that drastically vary in styles they produce. "AI" art can look very different, and I still think the only common property to it is sloppiness, lack of intentional attention to the fine details.
It's just not good enough. AI art generators aren't used in porn industry much. That should tell enough about quality and value of those. Bad VHS tapes were worse than YouTube 144p, and still totally fine for that industry. We're almost as far away from viability of AI art as we are from reliable self driving cars.
AI code generation is simply good enough. It might spit out GPL or Classified codes, but a lot of code just works. So they're widely accepted. AI translations and text interpretations are good enough too. READMEs are okay but barely. People are quick to point out Claude isms and reach for dislikes. "That's not just a double standard - it is a concrete indication that code and text are not the same" or whatever.
Have you been in a self-driving car?
Self driving cars are pretty fucking good, that's not doing the work you think it's doing.
I think we should be making games with AI however. I just don't think art is special; we have all been robbed and now we should all use the new tools that are the result of that robbery.
AI art, on the other hand, generally looks like obvious AI shit, and the user will see that and notice it.
Even human-made art will negatively affect the user’s experience if it just isn’t that good.
From what I’ve seen of AI art, nothing is yet coming close to the overall quality and effect a talented team of artists can achieve.
I think AI disclosures are good though for a different reason: when there are so many games, I want to know if this game was worth someone’s time and effort and money to make, or is just AI slop.
Also - the playerbase is sometimes ok with buggy games; for example - the first ARK game was a buggy nightmare, but it was still awesome and fun (it used the unreal engine visual development system...)
If you replaced your art team with an AI I won’t give you my money. If you lie about it then you’re just a scumbag. If your professional artists are using AI for their own work à la how I use it to generate code, I’m less bothered, particularly if they’re iterating on their own artistic inputs.
It’s just a matter of principle. Some of the evilest people on earth are the ones who own, operate, and directly benefit from the commercial proliferation of these tools. They commit crimes against us but see no consequence for which national-level politicians they’ve been sure to own. They’re aggressive in how they buy and corrupt local governments against the wishes of their constituents and the burden of responsibility is diffused because it’s not Anthropic buying your city council, it’s some other corp who focuses on building and operating power sinks that ruin entire neighborhoods and drive up their natural resource costs.
The reasons to hate AI (the industry as opposed to the technology) are impossible to ignore unless you keep your head buried firmly in the sand.
If one’s moral compass suggests that building a business on deceit is okay then I suspect they don’t particularly care about all of the externalities anyway.
Does this mean reducing number of employees on art team to zero is bad?
>If your professional artists are using AI for their own work à la how I use it to generate code, I’m less bothered, particularly if they’re iterating on their own artistic inputs.
Does this mean reducing number of employees on art team to a number greater than one is acceptable?
>It’s just a matter of principle.
I am having trouble understanding the principle. If I interpreted correctly, then what is special about reducing number of employees to 0 rather than reducing them to some other amount.
you can produce art with ai, but not everything ai produces is art, etc... when you produce code with ai it is always code
Code is more opaque, an ugly unmaintainable mess can still create a decently functioning game experience from the player's pov. There's infamous examples of some very shitty code still resulting in games people adore.
Ultimately your point stands, if the product is genuinely good and appeals to people, they will like to buy it and play it no matter how the sausage was actually made. AI generation just ruins the signal though, it's tougher to find the jewels among the sea of slop and that'll affect how consumers judge your product.
For AI generated assets getting coherence in terms of art assets not even just say like character art design, but artistic coherence is much broader and hard to make explicit. What is it that makes the colors, texture, models and animations work together well? What is the relationship between having an item with a specific color and texture in a specific context in relation to all those other items? That's essentially the question of what good art is and it's far more intuitive you could attempt to formalize it but someone can just disagree with your formalization and you can't really say anything.
Getting art that is coherent with the rest of your work depends upon having someone who is able to internalize it and express it. I personally don't think AI will ever be capable of that however I do think we will likely end up in a place where a good artist could leverage AI to produce far more assets then they could have individually. It will depend on them feeding it a bunch of structure and criteria and evaluating what it does though.
How can an AI understand enough about your character and his role in the story such that his personality is conveyed through his animation, his gait, his voice acting, the materials and clothing, etc. It's just not a strict syntactic relationship which is what AI is actually good at, there is no way to actually formally represent that.
The issue with AI generated assets for me is that they are just bad and ugly and being used by people with bad taste. I can see a future where you have technically competent artists having AI tools they can train or set up to reflect their vision and be closer to their work following principles they establish but we aren't there now. For coding because it lacks that a well written program by a human or AI (guided by a human) is basically identical. I could say an AI designed game would probably be awful for similar reasons that AI is or purely vibecoded things that are artistic like shaders, specific mechanical tuning, etc for the same reason AI art is bad. Those are also much easier to tune because they are at the end of the day symbolic formal representations that can be easily manipulated by AI.