Is AI killing the 3D star?
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
As Ai takes over more and more tasks, and fewer and fewer people have the skills to perform those tasks we inch closer to the precipice of scenerios like "The Machine Stops" ( E.M Forster ) and other similar stories, where one day the AI breaks or refuses to perform some task, and there is no human with the skills to perform the task either.
I think that's a much more likely way that AI will wipe out much of the population than it becoming malicious. It won't try to kill us, it will just not try to keep us alive.
Or that's just what they want you to think. "Hey, look my real family look like Muppets. How unusual, right?"
Remember the Eloi!
Exactly right. Our ability to think, learn and overcome complex obstacles is the reason we're survived this long. With no more thinking, no more trying to learn things, develop skills and our dependency on automation, it paints a bleak future. I get a huge sense of satisfaction when I overcome something difficult in my renders through work and finding solutions. The artistic eye and human perserverance is invaluable.
I know I've shared some of my AI stories here and there on these forums. I don't hate AI - I extensively use AI chat as a writing buddy, for example. Not to write the story - which is my job - but to chat incessantly about my characters or setting or plot or whatever I feel like blathering about today, which no actual person in my real life could even be paid to do, much less do so with willing enthusiasm. I also actually subscribed to Suno lately, as my "make music button" hobby had gotten to a point of justifying that. I do AI art at one of the many purveyors of such things for two purposes: 1) I create incidental objects like photos and paintings for Daz scenes and 2) I create images for my Suno songs. I do the latter because yeah, it's a lot faster creating a cool one-off image with AI than doing it with Daz. But to this day, I'd still rather do a planned image, featuring any object, setting or figure that I ever intend to use again, with Daz than fight with the AI. I want two wings on my eagle, thanks, not three. (Not joking - that was a fight for an AI image I did for a Memorial Day song.) I want four fingers with a thumb, oriented correctly, on both hands, dammit. (Also, a recent fight with AI for a Suno song image.) I want four legs on my horse. Moreoever, I want the horse to have the same white markings each and every time, from every angle, if I use that same horse in multiple images. In Daz that is instant and easy. WIth AI ... not that I've seen, and I don't feel like investing what little free time I have to argue with AI over things you take for granted with Daz.
I've told the story of an image I was working on with people inside a bar in which I thought to myself, oh, hey, I'll just have the AI come up with a basic street scene outside of the window in the background of this bar setting rather than go to the trouble of loading one of the various Daz streets I have. Something like 20 minutes later, I gave up entirely, having wasted days' worth of credits, because the AI would not under any circumstance create ... a street scene as seen through a window. It would only create scenes at intersections. Thanks to my IRL life, I have so little time and patience for BS that that ended my evening of trying to have fun. The Daz part was fun. The AI part - nope. That's why I carry that story like baggage. The fact that I still have to argue about things like wings and thumbs tells me that trying to create consistent images in AI would be an impossible exercise in time and patience that I don't have and wouldn't give if I did, since there are 15 other things I'd rather do with those resources.
I have a friend who is really into using AI for Daz postwork and gets great results. That said, that person is constantly working on detailed processes for doing it, and no small part of that work is about the preservation of consistent figures for storytelling purposes. I don't have that kind of time or inclination. I've been struggling a bit lately with being happy with my art work, but I still believe my goals are to create the image that I want, leaving time and energy to write the (%#$#$#) book, even if I'm sacrificing some perfection in the image.
It seems the disconnect between those who are fond of ai and those who are not is that the former believe that the goal of the process is an image, story, or song, and the latter perceive that the goal of the process is to be a creator.
Yes, for some people it's the journey that matters, not the destination.
Once we've arrived, we all come to that conclusion, even if we aren't immediately conscious of it. Eye has not seen, nor ear heard, nor has it entered into the mind of man and all that.
All the wannabe artists without any skills will switch to AI.
Real art created by humans will be harder to find and become more expensive.
I've seen incredible work done soley with AI by artists with decades of experience in 3D. Thankfully, 3D art is still in relatively high demand.
My creative output has been virtually non-existant since January because I've spent more time trying to get AI to p roduce what I want. Consequently, I recently found myself debating whether to put an end to my obsesive struggle with the tech. I started a poll last Wednesday, on two adult art forums I frequent, asking viewers if they would like to see more AI art from me, or if they prefer my 3D art. The poll ends this afternoon, but so far the combined results are 9 to 4 in favor of my 3D art.
Thanks to some great advice from another artist, I've already decided to press Pause on AI and re-focus on 3D until AI reaches a level of maturity that's better suited for my needs.
3D artists can have a better idea of what to instruct AI to make for them, but it still pales in comparison to what an artist (with skills) can actually produce. The only advantage AI has over those will the skills to create is time. And even still, it's debatabe how much time someone can save when the machine doesn't produce what you exactly want, and you could have done it yourself in a comparable amount of time.
My experience with AI has inspired me to just not even bother with it anymore. The time and effort spent with it is better suited to develop any practical skill in any other avenue. And as @alienarea eluded to, those who are able to perform tasks without needing AI will become increasingly valuable due to their rarity and the ability to stand out from all the AI imagery out there.
In my country the government will create a new gigantic AI center "to be ahead in the global AI competition". The downside is the amount of power it will use, which compares to the power usage of 2 million households, in a country with about 3 million households. Google, Apple, Microsoft etc.. already have huge data centers here which are using enourmous amounts of power and they will also use more and more in the future.
It would be interesting to see what would happen if an 'visual art' Ai was trained only on paintings created up to a certain date ( for example 1800 ). Would it ever progress beyond the styles of art it had been trained on and create art like the real artists did after 1800? I suspect not. In the same way using an Ai now is very unlikely to come up with anything that isn't a variation on what it was taught, it can't progress, it can't create 'original' art. If all artists were to switch to AI then that would effectively be the end of progress in visual art.
that's already a problem because the popular method of training LLMs is to scrape everything visible on the internet, and it turns out if you feed it its own output (or other LLM output) it poisons itself. pretty cool!
https://www.nature.com/articles/s41586-024-07566-y
the great thing is the taint is totally intractible, whatever gets in, you can never get it out lol
A good friend of mine once told me that "Any digital system can be hacked, And because it can be, it's only ever a matter of time before it will be.".
He's a professional systems penetration tester and ethical hacker, so I guess he's right, and if so then that statement must also apply to A.I.
i'd really like to see their workflow.
My understanding of an AI workflow is to create a prompt and then select the outcome.
Obviously, the AI- "artist's" skill is in the ability to create a text, that describes the imagination, and then making a choice.
On my side, I cannot tell you how often I reset lights or a camera angle about a few degrees until it looks right to me. Struggeling hard to get that perfect pose for that moment.
Do the characters in the scene have the correct distance to each other? Are the expressions and their focus right?
And no, human creators don't just copy like AI does, they put a diffuse mix of personal experiences into their art. (just in case, that argument rise up again)
Closing this as it was getting heated and politics-adjacent.