I understand the need to place something in your game while it’s in development. But the moment it touches the public’s eye, any AI assets need to be replaced with real art.
Companies have placed temporary assets into their games for decades without genAI. I find it very, very hard to believe that the temporary genAI assets are more valuable than the previously used solutions by enough to justify raping our planet’s natural resources and stealing the sum total of human knowledge and creativity for profit by like five guys.
You’re kinda alluding to it yourself, your issue is with the abstract costs.
As a software dev, I hate “AI”. I actually generally hate technology, period, but I digress.
That being said, lofi temporary stand-in assets are pretty much the ideal use case for generative ai. It let’s you extremely quickly iterate on ideas. Think that scene from the Simpsons poochie episode where they’re designing poochie… “try making him 10% more rastafarian” “no, too much”.
They’ll all be trash, but it would let you dial in on feel in minutes instead of weeks/months, just because the loop is so tight. It’s an objectively a good tool for that specific job.
Your issue is with the ethical costs. The environment. Plagiarism. Totally valid.
But you have to recognize that no industry is going to thrown down thier tools willingly for abstract ethical concerns.
This is why regulation is critical. If the tools were objectively not good for the task, I wouldn’t care about regulation because the problem will solve itself: if it’s worse than the “old way”, people will just go back to the old way.
But for this use case? It’s better. It won’t fix itself. Nobody is going to grow a conscience. Without legislation this will absolutely be the new normal.
I’d love regulation and legislation.
I agree, and wasn’t attempting to justify the use of generative AI (I myself am vehemently against it). I’m not sure what the solution is here besides to stop using it; if I were a game dev, I may find myself tempted to use it because I have little artistic skill, if I had no other resources for art. On the other hand, I’d boot up MS Paint and start crudely drawing my assets until I got a real artist, but I suppose that’s hardly presentable.
If they told us that they used a locally hosted model to generate assets based on their own art, that would be perfectly acceptable. But we know that’s not happening.
I wish everyone else would push back against AI as much as the gaming community pushes back against it.
Upset they got caught, more like.
Oversight? Oversight?
They used AI to generate the apology.
It’s not a lie! It’s just creative use of language!
It’s a demo for an unreleased game. Give them some slack. Placeholder art is nothing new. Using AI to generate it is a legit practice because you aren’t wasting an artist’s time on shit that won’t make it into the game.
Placeholder art is nothing new. Using AI to generate it is a legit practice
Funny how placeholder art existed long before AI slop, and it didn’t seem problematic.
Using AI to generate it is a legit practice because you aren’t wasting an artist’s time on shit that won’t make it into the game.
I’m sure those artist’s would like to earn a paycheque regardless of whether their work is in the final game or not. That’s my main beef with A.I.
Edited: The amount of slack I’m prepared to give in the case of placeholder assets depends entirely on the size and financial backing of the development team creating the game. Am I going to flip out about a small developer with very little money creating their small passion project for using AI because he can’t afford a bigger dev team. No. I’ll wish him/her all the success in the world so that they reach a point where they no longer HAVE to use A.I. to generate placeholder assets. But will I give a billion dollar company slack for using A.I. instead of paying artists? Hell no.
Long story short. If you’re a big company with the money in your couch cushions to pay artists, than PAY ARTISTS. If your a small indy dev just starting out, A.I. placeholders can help you a bit until you can afford proper artists.
Every single generative model in existence was produced by stealing the work of artists for training data.
There is no nuance to this. Never mind the argument about generative AI taking away potential future work from artists. The training of these models is theft now. To use these models is to be complicit in the theft.
Edit: the people who downvote this while making no counterargument are doing so because they wish to remain in denial, so that they do not have to admit to themselves that they are complicit in the theft.
Not to mention the amount of water and compute required to generate those temporary assets. Meaning they are complicit in killing the planet for something they don’t even intend to continue using.
That edit: yuuup 🎯
Many game developers would disagree
For many it isn’t a waste of time, it’s part of the process of exploring ideas to find the best ones. A classic example is Nintendo’s Kirby he was placeholder art they decided was perfect from the start. You wouldn’t get that just shoving in AI garbage. A more recent example is all the placeholder art in the Slay the Spire 2 early access that everyone is in love with.
So no, no cutting them slack for being lazy and dumb.
This is a valid point
But so is the perspective of a single indie dev using ai placeholder art to build a prototype with the idea to attract or hire an artist when the concept has matured into an actual plausible future release.
Except they don’t need AI to do that. That’s literally the point of many of the experienced developers who have been sharing their placeholder art.
You are confusing correlation with causation. Placeholder art doesn’t give great art assets. Artists do. Just because there have been a few times in history where placeholder art got it right doesn’t make it worth preserving in its current state. Now, instead of spending time on placeholder art, the artists can do what they want to do and be creative with designs. The doodle that became Kirby will still be there. The creativity isn’t going anywhere.
Not confusing the two. Exploring their ideas and seeing how even the simple placeholder art looks can spark more ideas for them. Placeholder art didn’t give great art, no, but flexing their creativity and exploring what the game will be can. It doesn’t cause it but it’s sure part of the process. It’s why concept artists are a thing and their entire job is just toying around with art that will never make it into the game.
Think I’m gonna trust the people making beloved games, not some Internet slop apologists on this. And time and time again artists say no to AI and would rather be allowed to cook.
“Placeholder art doesn’t give great art assets. Artists do”. So let them do their thing, not AI.
I disagree. Placeholder art should clearly be placeholder art. A stick figure made in Paint is a better placeholder than AI slop. It makes reviewing easier too. Plus you’re not burning tokens, compute and the environment for something that won’t make it into the game.
Gamers are clearly anti GenAI for assets. Any studio who uses GenAI for assets, even placeholders, is taking an easily avoidable risk. Players will find out.
“…wasting an artist’s time…” doing his work he gets paid for?
I somewhat agree with you but not in full. I agree that AI might be great for generating placeholder assets because it allows someone without the artistic skills to produce images/assets that convey a certain idea better then trying to explain it with words. A picture says more than a thousand words after all.
However, I do think that AI assets have no place in a product delivered to consumers. Not as a demo and certainly not as a final product. This goes for any industry and any type of AI generated output. It should merely be used as something that supplements a written idea with another format so to adequately transfer ideas from one person to another.
unfortunately, lemmy tends to be extremely anti-ai anything. so you will rarely see that kind of nuance here.
but I agree with you.
edit: see? told ya. lol.
Every single generative model in existence was produced by stealing the work of artists for training data.
There is no nuance to this. Never mind the argument about generative AI taking away potential future work from artists. The training of these models is theft now. To use these models is to be complicit in the theft.
Edit: the people who downvote this while making no counterargument are doing so because they wish to remain in denial, so that they do not have to admit to themselves that they are complicit in the theft.
BeYou are the change i want to seein the worldon lemmyNot ai singularity cult, not modern luddites. Actual nuance.
welcome to the downvotes, m’friend. 🫂 we chill here sometimes.
