AI is going to be our biggest game changer
This discussion has been closed.
Daz 3D is part of 
Connect
DAZ Productions, Inc.
7533 S Center View Ct #4664
West Jordan, UT 84084
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2025 Daz Productions Inc. All Rights Reserved.
Comments
I keep getting stuck on words. To me using a machine learning image generator is analogous to hiring a commissioned artist. Your imagination. Your description. Someone elses creativity and talent. Except with the machine learning image generator many of the actual human artists didn't have any say in whether their work could be used and whether they get compensation.
Alot of the output shows the image generators screw up simple things (like limbs, hands, heads, etc) as well as sometimes having features on a building where the perspective isn't consistent, and other mistakes that imply there's no "understanding" (yeah, not the right word) of spacial relationships between objects in the image. There simply isn't any intelligence here. Just math.
It's still early days for this. When they train in 3-dimensions instead of 2 and have an "understanding" of what a house, road, person, hat, etc could be then it would be closer to Artificial Intelligence art. Until then it's just a marketing buzz-word. One that will be useful to many people, but oh so far from realizing it's potential for creation.
Does that output then become open source in terms of copyright? Because everything created with Stable Diffusion is listed as open source, correct? That's my worry...run my own pieces through an AI and lose copyright. nothankyou.
Except now the artist isn't getting paid. That's one of the big things about this whole thing...artists could end up out of work because people can just enter prompts to get the fanart they're looking for...or whatever art they'd commission for whatever reason. The artist is now out of the equation.
Yes. I agree. That's why the rest of my statement was: "Except with the machine learning image generator many of the actual human artists didn't have any say in whether their work could be used and whether they get compensation."
If original, human-made art and illustration goes away how will the AIs be trained further? It sounds like a good recipe for stagnation, which would then reopen the field for humans.
The AI needs human input to have materials for kitbashing things with. Not unlike using Daz or Poser. It also offers no complete control over every detail of the picture. Not unlike creating with Daz or Poser.
Ironically, I'm not concerned with the AI-generated generic art because it was exactly lack of fine control over all aspects of the image what made me start moving to trying to draw things by hand.
Just wonder, how this AI works for creating art.
Is it only trained on good looking art, or maybe also on the other, not so successful ones.
In chemistry AI, it is trained on both succesful and failed cases,
to learn how to avoid make mistakes in the future.
I don't believe human-made art / illustration will every go away. Many people have a need for expression for a variety of reasons and they will continue to try to fulfill that need. However, as I understand it, the output of the machine learning generated images also gets fed back into the system somehow. It's not totally dependant on humans creating new art to feed it. It certainly sounds like it could become rather "inbred". :-) Also, there's alot of human created art out there that hasn't been fed into the systems yet. Wasn't there an aticle recently about one of the sotck image vendors signing a deal with one or more of the "AI Art" companies? They could add a new clause to the contract that allows the company to use newly submited art for machine learning, or the company could pull a "All your stock images belonging to us. But don't worry we'll pay you a penny, once, for each image of yours we use no matter how many times it's used to generate a new image". I wonder which they'll do.
"AI art" generators aren't a tool for everyone. Just a tool to monetize people who like the new toy and aren't worried about what's in the booth behind the curtain. And that doesn't bother me. My Nikon digital SLRs have some kind of image processing algorithms developed based on Nikons huge library of images and that's a good use of the math as long as I can still get RAW images out. :-)
It does annoy me somewhat that there are people out there that don't see a difference between the work of people who invest time to learn technique and effort to develop talent versus what pops out when someone types NVIATWAS and gets an image. But I've been annoyed before and it hasn't killed me yet. I don't feel the need wipe out everything that annoys me. I check the midjourney showcase and the stable diffusion discord channel and some other website (KREA ?) almost daily and there are lots of pretty images there. There are also a fair number of failures (wrong number of legs/arms/hands/etc) and a number of really horrific images that are probably intentional and just "that style" of art. I mean H.R. Giger made some genuine nightmare material and it worked great in the "Alien" franchise. Hmm. I have the box set. I think I know what I'm watching this evening.
I can see the day when people talk to their TV and wind up getting bespoke art work, other home or car deco, or clothing. All one offs and each as unique as the person that spoke the order cared make it. Maybe, they didn't care to make it unique and instead ask for a tux or dress from 1920 (to remove copyright / trademark problems)? Computer AI is getting very close to making it easy for a lay person to feed a few images of themself to a computer and get a realistical 3D model on which such programs can use as a demostrative clothing horse.
I wonder how much fuel and other raw materials are wasted every year on manufacturing clothing that never gets bought? Many more physical products are similarly speculative in consumer demand every year.
I am no expert on “machine learning”
but it would seem that the millions of AI generated images that already existing at such a point ,presumably multiplied into billions of variations that no single human would ever recognize its original source “inspiration”.
We have to assume that AI’s are not just using some “still “Art created by some famous Artist.
what about every frame, from every movie, cartoon,Video game cinematic,youtube video, already in existence, and still being made, after the still Artists become “extinct”.
Takes a very long time and a lot of aggravation to make 3D objects. It's like walking from London to L.A. threw mudd up to your neck while carrying a elephant.
a.i. can't get here fast enoght.
AI has taken root so strongly that vendors are now selling prompts.
someone is selling the keywords as a packaged product.
Oooo! NFT the PROMPTS and *really* clean up!
DeviantArt site has an AI generated art option called DreamUp which I tried out. As an artist there is something very uncomfortable with just adding prompts and voila....
Not sure how to feel about all this. Change is coming whether we're ready for it or not.
Taking something for free and then try to commercialize it? Would you pay $8-$12 per month for a AI generator?
I know how to feel about it. It can't be said here.
Get some understanding about this AI nonsense.
After giving it some thought, I think I'm just going to get the 4060 or whatever comes with 12Gb. Having a 1080ti was nice, but if the 60 level card offers better performance and more VRAM, I'll go for it.
I mostly play older games which will never get DLSS, so I'm using FSR for upscaling.
It sounds like the reaction from reviewers has been mixed. DLSS 3 (and FSR 3) use some type of frame generation to create frames inbetween rendered frames and it seems to be causing artifacts.
Thank you for the link to this,
bottomline, "Artists should be mad about this ! "
Cheers to you! :)
interesting what is trending on Artstation
It's too bad that visual artists wouldn't get the same protections that musicians would in the event someone creates/releases a Stable Diff for music. Radiohead can sue if your song sounds a lot like theirs but Greg Rutkowski can't sue you if you copy his work to the point of it being indistinguishable from his works (down to his signature) and then sell it as your own.
Its a slippery slope - some people think the same of using Daz Studio or any computer generated art for that matter
https://twitter.com/Vanessid/status/1600563042302861312
Going that route you could say the same about photographers, though I would disagree with both.
It was said by photographers at the time :) I was myself a bit annoyed at that point in time, having just finished photography as part of my degree. I wouldn't give up the ability to make digital photos now though - I could document my kids growing up without paying an arm and a leg to develop photos.
edited to add - working with photographic chemicals is awful - I won't miss that either :)
That's the thing, there are plenty of photographers I can point to and say "that is a fantastic artist" because they have an eye for composition, contrast, and ways of capturing a subject that is unigue to their style. They have an artistic identity, same goes for some of the folks here. Same cannot be said of people who use prompts. Though I guess it would be another discussion when it becomes to photobashing with AI output.
I don't disagree with you - I'm mainly pointing out that its super easy to turn the discussion from AI, to CGI and any kind of digital art... especially if the users are not using any computers at all - case in point the twitter discussion I linked to. Its not the only one out there either.
I do however agree with the title of this post - it is going to be a game changer - not just the art side of it - but all the other AIs training at the same time on data sets... and as an artist I am making very sure that I'm up to date with the latest developments, including on how to use it. The genie is out of the bottle.
Art. Community. Surprise. Love it.
https://www.artstation.com/?sort_by=trending