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
AI is not killing it, but becoming part of the workflow, part of every software, and hardware. (AI alone is weak and repeats itself, but to some people that is better than what most artists do. This tells about us as humanity. )
If you think about it with styles, looks, actions, generative fill, etc.. it started even before, but many people didn't notice it. You put an image into a software and get back 100-200-400+ alternatives, different software handles different needs. Do you need any technical knowledge for it? Barely. Do you need any artistic knowledge for it? Barely.
In games, there is already AI between frame generation, the next step is rendering outside of games. (5-8-10 sec AI clip generators are already out there) Soon raytracing / Iray will be in the past too. You need a text & 3d bare idea and the AI will render it for you in 5 minutes. How many nights, and hours have we been waiting for the renders? So I'm not missing the old times!
Facts ^
Our civilization will not abandon this technology any more than it will abandon digital photography to regress back to chemical and paper photographic prints.
Also any notions about “corrupting the training data with “bad Art” no only lays bare a woeful misunderstanding of the tech in general, but it assumes that all “Artists” are a monolith
who would waste their time on such a failed effort or would universally refuse $$monetary$$
compensation for their Data( were it offered) and remain poor and broke for “The cause”
Finally for those who make the “AI art has No soul” argument…well ChatGPT
Disagrees.
I rarely agree with Your posts, but on this one I can only fully agree with it and Your hopes.
I am willing to consider AI in its various forms as a collaborative tool. In terms of art, I am waiting for AI-as-postwork to finesse an image created by me in Daz: a "Make It Better" tool, not a "Make Art" button. I would not have an ethical issue with using AI in that way to make small fixes/improvements on an image that I created any more than I would have an issue with using Photoshop to do postwork. At the moment, it doesn't seem to be there. I have a friend who was doing a lot of work on that front and can get impressive results, but for my taste, it's still modifying the original image too much. But when it comes - and it will come - I will add it as a tool to the toolbox.
Outside of art, I use chat a great deal with my writing projects. Not to write my stories for me, because I'm the writer, and that's my job. But I have found that an AI chat - I like Claude, YMMV - is the best writing collaboration buddy, ever. No one I know IRL has the remotest interest in my fiction work, but the AI chat will gladly spend all day enthusiastically chatting about any nuts-and-bolts writing topic that might come up. I use it as a sounding board and in the process of writing about the topic, I often come up with ideas that I might not have, had I not been able to have a conversation - as it were - with somebody/something.
A person can copy the style of Monet so exactly that a museum would not be able to tell it wasn't painted by Monet. I've read in considerable depth about art fraud (maybe 10 books), and if you think a human can't copy a painted masterpiece by virtually any famous painter, you're wrong. And they can copy their style - so they can paint original art works in the style of Monet so well, that you can't tell it's not a Monet. Museums need to use advanced technology and science to analyze paintings and sculptures to tell if they are authentic. The greatest art experts in the world can't tell if a Rembrant was painted by Rembrant, or by some guy in his garage in California - just by looking at the painting. So it's not just AI that is stealing "real" artists work to copy to create new art, humans are doing it without computers - and they always have.
this the problem with the AI debate
the all bad v all good dicotomy
some try to address the ethical training issues only to be told it's giving big corporations like Getty Images using their licensed content leading to a corporate monopoly
or it's soulless, taking jobs etc
others like to push the democratising aspects saying it gives unskilled people an outlet for creativity like using 3D models posed in DAZ studio does to an extent and we all know the hate directed by pure artists at doing that
there are far to many layers to address and have a single debate
photogrammetry is probably the bigger threat to the 3D star specifically
AI + photogrammetry + 3D mesh creation from single images with AI assisted retopology will be the nail in the coffin for 3D modellers
I don't see that as a bad thing, it's a tool like the camera which was the first step towards capturing reality
the skill will always be storytelling, composition etc, having a phone camera doesn't make one Dean Semmler
enter ChatGPT and other large language models...
I created this yesterday with AI. It's MY creation. I had to write almost a hundred prompts to get it to look like this. I can create 8000 different images a day using AI, and I often do. AI is way way way beyond anything a human can do alone. It's advanced past perfection. (don't comment on my "taste" - I made the room look this garishly grotesque for a purpose.)
I am a software engineer, and so my job is just as at-risk as anyone's (whether AI can replace an engineer or not is a moot question if hiring managers already think it can).
That being the case, I am still more excited about the new possibilities and modes of creating things than I am anxious about the old model being destroyed.
I think the trick is to be open minded about how to ensure that AI empowers you instead of replacing you. And this is not different from any new technology since, say, the Renaissance.
I mean, just look at the links WendyLuvsCatz posted. Can you not sense that the authors has a vision that AI realized for them, probably in a way that would have been impossible or impractical without it? How is that different from technologies from the invention of pigments all the way up to something like DAZ Studio?
All this talk of it being soulless, unethical, self destructive, inbred, is just because AI is obsoleting a higher layer, a layer that people are attached to for myriad reasons, from AI replacing the part that is enjoyable to them to AI making it more challenging to earn a living doing something that they love.
AI is no different from any other technology: You're either empowered by it to create things you couldn't have before, or you're replaced by it. This from a guy whose job it's going to destroy; the consolation is that I'm doing the things I want to do better and faster, and even things that I would not have even attempted before: mocap, soundtracks, voice overs, rotoscoping. All that was inaccessible by me before AI.
The world is partitioned into to types of people: those who, like all those people before them who no one remembers anymore that resisted the inevitable, and those who saw the change and figured out how to master it instead of it mastering them, i.e. the harder approach.
This is a false equivalence.
Yes, talented forgers have been able to produce indistinguishable imitations since the beginning, but the ability to do so has long been fairly rare and required long and skillful study - and yet it was still enough of a concern that the museums still need experts, technology and science to sift out the imitations. (Whether or not it's valid to value a genuine Rembrandt over an equally skilled artist in the same style is another argument, but at the very least we can acknowledge that scarcity creates value, and so the markets consider the work of one Monet to outvalue the work of a thousand imitators).
Now? Eh, throw a few dozen images on to a GPU cluster for a few hours, teach it to churn out ten thousand Van Goghs a day.
That it was a manageable concern before (and even a legitimately marketable skill for the artists in question; you don't necessarily have to be trying to create forgeries for "I can paint like Renoir" to be a skill people might pay for) does not mean that the capacity for creation of imitations exploding a million-fold is not devastating to the market.
Do you look forward to a time when the AI will do thatfor you?
I didn't say everyone in the world could forge paintings - I said it has beeen, and is being done by humans. And your thousands of AI Van Goghs, Rembrants, and Monets, are not worth anything.
The AI can already do it for me. You can train it to give your own personal style to every image. I imagine that you could tell it to only produce images of cats. Then every cat image will be in your own personal style.
This is something that no human can do. Only AI can do this. This is the face of Sean Connery and Montgomery Clift blended into one face. (I made this image with AI)
I've never heard of photogrammerty - but whatever it means - only YOU are going to be able to create your characters and your creatures - that exist only in your head. So it will be like everything else in the world - people will pay a premium for sometheing they really really like, something unique, and you can create that, and AI can't.
photogrammetry is what Quixel Megascans among others do, Dreamlight sells many examples in the DAZ store as does polygonal Minatures
it is basically taking lots of photpgraphs or video footage and joining it all together to create a point cloud (a lot of points in 3D space)
3D scanning is another name for it
the point cloud can then be made into either a guassian splat or actual mesh using various softwares to create faces between the points
one can then use retopology such as Zremesher in Zbrush to create clean meshes that can be UV mapped and Zproject the vertex colours from the original on to that
3DZepher is a photogrammetry software you can use on your own computer as is Meshroom
But it's not as easy as typing a sentence and clicking ENTER. These 3D models that you say are alreay being created and sold here - are they rigged, or rigable? Or are they just a static mesh?
can be either
DAZ uses Photogrammetry too, nothing new about it
there is a video of Reby Sky being 3D scanned for V4,
it's just that now one can use any camera including a phone because of software advances
partially aided by AI
the term AI does not just mean GENERATIVE AI BTW
or Diffusion models
Face Transfer plugin is such an AI assisted tool too
Pose Recorder too
The image used for that was cropped, there was nothing salacious about it. As for commercial AI being trained on pornography, if it is, it is blocked well. When I used Midjourney, I got warnings for trying to enter "hip" and "cleavage" in my prompt. With CivitAI models, I prompted for "water pool in a cave" and got a nude woman in it on 29 out of 30 attempts (yes, really), even when putting woman, female, girl, etc in the negative prompt.
Most the time I want the output to be something that feels fundamentally the same as the original. Here I pushed things a bit farther. We have a reimagining of the K4 Alien Force (Kids 4 was a figure from the same generation as Victoria 4, pre genesis, back when Poser was actually the dominant bit of software).
Then we've got a couple of my other personal faves from over the years, the Genesis 8 Snowman and Genesis 8 Dragon Hatchling.
I loved that Hatchling version more than Drago himself.
Though, really, there was something to be said for that old Alien Force. I think they were made over the course of about a single snow day off from school. Imagine a time when you could just bang out some interesting head shapes in zbrush, throw a procedural texture made in the poser materials room on it, have no end of fun cranking out endless promos and call it a day (This particular example was a Visual Style render from Daz Studio, though).
Absolute night and day. Were those done with a Pony model too?
I think so. I'm druging up stuff done over a span of time here.
As Illustrious moves beyond just anime, I'm starting to play more with those models. Since it and pony are built off of sdxl, a lot of LORA seem to transfer over alright. Illustrious seems to have a broader understanding of the universe in place of all of Pony's kinky specificity. It seems easier to prompt and steer, if maybe a bit lower quality at other elements.
I also noticed that controlnets, if run in balanced vs prioritize prompt, had weird effects in Pony. Prioritizing prompt eleminated the weirdness, but it also diminished my control. Illustrious doesn't seem to have that problem.
I've got the power to run Flux locally and quickly on Forge (vs automatic1111). But I think I would need to learn ComyUI and it's nodes to actually use ControlNets.
Thank you, I don't mean to put you on the spot with these questions. I have not had a chance to try Illustrious yet, but I will try after hearing your thoughts on it. You're right, Pony can be tough to handle sometimes and it's trained on so much...questionable content. Browsing Pony LoRAs on CivitAI can give you nightmares. ;)
I wish I could get the results I want from Flux, but I just can't get its official controlnets (depth and canny) to read my own work at all. Even in Comfy, there are a dozen ways to set up Flux nodes and none of them have really given me great results. That Alien Force AI generation did such a great job of giving those very...blobby characters unique personalities and looks.
And the next step will be the AI delivering a picture you like without You even pressing a button, as it will be able to notice automatically when You wish to have a new picture and what it would be supposed to show, right?
Would be the ultimate goal for an AI user to get results without even having to do anything for reaching them. Of course human artists would become useless sooner or later, when AI would supply everyone with all pictures one could want to have automatically. Might also lead to a point, when those AI pictures are just a nuisance, like advertisement is for some already...
In this vein, I decided to pull out one of my older AI generations (admittedly I think this was SD 1.5, but I haven't liked any of my attempts to redo it with newer models)...
... and instead redo it properly...
(All skimpy outfits are improved by horse).
That's rather my point.
Things only have financial value because of scarcity. AI generated images are so plentiful and accessible as to be functionally worthless (even putting aside their generally inferior quality).
It seems there are at least two separate discussions going on.
There are the first group of people who view the AI as a way to create an entire project. And some of them think that is invalid because the AI must be ripping off a bunch of other projects and putting together pieces from each of them into a form just different enough to avoid copyright. And others of them think the project created by AI is valid because it can take whatever someone wants and make that, regardless of how it learned to make that; and the result is a realization of that person’s vision rather than the vision of the artists it was trained on.
And there are the second group of people who see AI as postwork for a project created by some other means. And that group has some people who think the AI improves things while others think it isn’t worth the trouble.
And there are some people who just like what they’ve been doing and don’t want to change. I don’t know.
Strange you say that because I've noticed that when partially nude women appear in AI imges on the "Explore New..." page of NightCafe.Studio they almost invariably are standing in water. Usually it's the torso but sometimes the behind, nothing really risky. I'm not sure how they do it though as I've not enough interest in trying to break their safeguards. I'm not even sure it's considered "not allowed" as it's OK for AI images of men to be topless.
Those are awesome refinements.
Count me in for the second group. As soon as there is an easy (aka not having to type hundreds more or less cryptic commands) way to enhance my works (without changing too much or coming up with unwanted changes) I'd be ready to try AI as a tool.
Right now, I don't expect that to happen soon, though...