disseration survey on ai for games art

tbenstead05

https://docs.google.com/forms/d/e/1FAIpQLSdlacbXMY8XdVs_nfOcQWQTa5icU1719-68URfyp61O25X0wQ/viewform?usp=header

im currently doing my final year project and for my research i have created a survey about the use of generative ai in the games industry
any responses would be much appreciated thanks 

pior

If I may ask ... isn't it somewhat of a very boring topic already ? There has to be more interesting things to cover for a final year project other than subjecting yourself to slop for months ...

tbenstead05

i find it quite interesting and my supervisor agrees that there is a lot that can be discussed wiht htis topic. my project is specifically about the benefits and effect of using it for games artists

pior

Well, I can see how professors would *love* generative AI since it allows them to display the illusion of skill in front of a room of students - as opposed to actually teching them something rooted in their credentials (if any). And that's also ignoring the fact that all implementations of generative AI so far are based on litteral theft.

But hey, you do you. Just saying that a final project consisting of, say, studying the level design principles of a great series of games, or diving into the game design philosophy of a successful game director/designer is likely going to be more valuable and teach you more things.

zetheros

there is a thread two threads down from this one https://polycount.com/discussion/231978/ai-art-good-or-bad-a-hopefully-nuanced-take-on-the-subject/p1

As for your question, I'll give my answer in the form of a hypothetical question; "Dear US Department of Commerce - hi, I was wondering what role and benefits there are of counterfeit money in the US economy?"

tbenstead05

i posted on this thread more to get general responses for my dissertation research to see people opinions 

kanga

The term AI is misleading when really it's machine filter theft. In a time of spectacular corporate greed (lookin' at you merca) with all the horrors associated. Posting this here,... son,..... they must be so big, its' a wonder you can walk.

Ashervisalis

You're coming to an art forum and asking artists what they think about generative AI art... You're brave, I'll give you that hahah

Eric Chadwick

Moving this from General Discussion over to Career & Education, where these kind of topics fit better.

Feel free to canvas our users for ideas about AI in game art. However, as stated, there is a long-running thread that is pretty clear in expressing the community's overall feeling about AI in game development.

Overall I think the sentiment can be summarized as:
* Gen AI steals artist's work. This is shitty.
* It sucks for generating quality content. (but it might be usable for helping people generate raw ideas, which then require a human mind to refine and grow into a usable prototype)
* It masquerades as a subject-matter expert (SME), but human SME's know it's actually more often full of shit than not.

Anyhow, good luck with your disseration. I think our members stated it pretty clearly. AI is just the current hot topic, but no substitute for learning real knowledge. 

poopipe

I've responded - because I choose to believe the OP is asking in good faith. 

I'll add some extra notes as well.  
Where I work we are expressly forbidden from using generative AI for any shippable content - this decision was made at the topmost level of management (nobody is arguing with that). This is certainly in part due to ethical concerns but I believe there is also a legal aspect to the decision. 

ignoring my personal feelings about the ethics and the damage that not learning for yourself does... 
The fact is, Generative AI is not actually very useful for art content generation
1: It is objectively incapable of coming up with anything novel
2: *if* you give a shit about the output you need to train your model on your own content - this process is not just extremely expensive and time consuming but it's also not very reliable and it's almost certainly a better business decision to just pay some people to do the work. 

The tech underneath is amazing and will eventually lead to really good tools that increase artist productivity but we have a few years to wait before the investment circle-jerk comes to it's natural end and the smart people are allowed to do something useful with their time.