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 don't know how to with theGUIs I use
I believe they're referring to either Textual Inversion or Dreambooth. Here's a comparison of the two https://www.reddit.com/r/StableDiffusion/comments/xqi1t4/textual_inversion_versus_dreambooth/
I think the takeaway is that Textual Inversion has lower system requirements and it's included in some Stable Diffusion UIs. Dreambooth seems to require more VRAM and is a separate install. Also, from the comments, it seems that Dreambooth can train with fewer input images and produces a more faithful reproduction. Textual Inversion is more 'creative' and produces more variations on the original.
But definitely don't take my word for it. I guess 'faithfull' could mean 'boring' or 'creative' could mean 'you ruined it!', depending on what you're looking for. But that's art.
Thanks Wendy and Kitsumo, I will take a look at that video and link shortly.
Actually getting a little tired of hearing the AI-art hate lately because a) it is more difficult to get the exact results you're looking for than the critics think and b) DAZ Studio, Poser, and other pre-made content users should not exactly be the first in line to complain about being able to make art with little talent or effort. If 2D artists and those who can sculpt a figure from a sphere in Zbrush have a problem with it, I'm all ears.
The enabling part certainly is highly interesting, no doubt. Yet, there is more meat on the table than just "ai haters". E.g. copyright nuking, future cloud services diverting everything into the hands of a few investors, while their cloud beast feasts on everything that has been published (hell-version), which would include copyrighted work both by pa/artists as well as that of "premade content users". But that's probably not the "critique" you're getting tired of, rather.
b) DAZ Studio, Poser, and other pre-made content users should not exactly be the first in line to complain about being able to make art with little talent or effort.
Yeah, this comment struck me as a little off becuase it's the exact opposite.
Daz Studio (Asset users) are given LITTLE RESPECT by those who DO NOT KNOW what the Daz user does and does not do - to make their works/art. Once someone uses Daz or understands what it takes to hit the actual "art" level of a result, it's impressive. Hence the level you "art" has to be to get likes in any Daz-user group.
AI is reversed. People who do NOT KNOW what AI is or how it works GIVE TONS of ARTISTIC CREDIT to the programmer until they find out what AI is and how it works.
--------
What's happening is, there's an Assumption as to how these tools are being used.
There are also the environmental implications with all that machine learning. Makes me wonder how much of a difference there is to that level of computing and crypto-mining.
cannot argue after getting my electricity bill
felt dirty, had to spend some of my emergency funds
I have reduced my usage of machine learning apps a lot since spending most of my time editing the results doing 2D animation with postwork
video
I had the impression that many consumers of the current machine learning text-to-image generators were giving "TONS" of artistic credit to themselves rather than the specific platform that was piecing together bits of other people's work; but I could be wrong.
I assume the consumer is using the tools to (in no particular order):
- replace commisioned art work
- just have fun wasting time to see what pops out
- get new ideas for their own work or practice material to learn to duplicate
- create avatars
- make desktop images
- replace commercial sources of stock images
- create source material to use with other digital editing software
I am somewhat curious about what other uses consumers might have in mind for these text-to-image generators.
I assume the tool providers are using the tools to make money off other people's work without the artist having any say on compensation.
I don't really follow what people are up to (I'm a grumpy old-ish retired person) so I'm kinda dim on the subject but I'm wondering what assumption you were refering to.
Haha,
environmental... difficult to judge mid-term
- Specialized "fuzzy" but terribly efficient hardware?
- Cloud services rather, due to model size, IP and lock in factor?
- Copyright or not?
- Copyright too difficult? Maybe they just use this as a showcase, and there will be tons of other (rather more specialized) applications, with no risk of copyright fraud (due to xyz measures / data sets / etc.).
- Edit: Killer argument: You save so much time, so even with that bill, you're .. uh... saving time? Or everybody is now faster and better, so everybody now has that bill, and...uh... we're just at ... something like fracking?
Concerning the pre-made part - in a way there is a similarity, but the difference is, that with DAZ you pay the creators for using their works for your "collage" (most of the time, in any case you're using their works having their explicit prefetched consent). So on the one hand, bashing on the technique per se... ok, you can make an argument about "pre-made" there for sure. The classic painter could start at zero (though not all do). On the other hand there are clean cut differences, as mentioned, and indeed it should make sense to distinguish, what the actual aim/base/concern of critique around this topic is.
OK, a few things.
No disagreements there, I am only referring to artists using canned content complaining about how AI isn't art and is just 'cheating.'
We must live in opposite worlds then, because this is also the exact opposite. Most people who know nothing about DAZ products and 3D software assume that the artist created everything in the scene. It's not until they see that everything was made by someone else and just arranged that they might start thinking that 3D art isn't as challenging as it can be.
I'm not taking sides against artists like us, I've been doing this for 23 years and I know how difficult it can be to make a pleasing scene even with purchased content. However, what we do as asset purchasers and users (not the actual content creators) is only a few steps above AI image generation when it comes to the work required to create art. Concern yourself with copyright usage, environmental impact, and those selling AI art as stock images and NFTs, but let's not be too quick to criticize AI artists when we can buy entire environments with lighting and textures already set up for us, drop in a fully textured figure, apply both an expression and a pose to it with a double-click, and call it art.
There is frequently quite a lot of work required to make ready-made content look good in a scene. There is a lot to sort through There is quite a lot of knowledge of 3D techniques and art needed to get good results and it has long fascinated me that so many people are willing to do that work just as a hobby. The other end is, it's easy to become a content collector, to have more content than can ever be used. I'm impressed with people who get good results in Daz. I'm repulsed by most everyone using AI generators to make art.
"most everyone", eh? So ... got any examples that aren't repulsive? Or perhaps, less repulsive, in your opinion -- I will admit there's much incredible junk out there but then again that's true of practically everything, not just AI art.
No. I have no examples. I am thinking more about people. I have been friendly with some people who have whole-heartedly embraced the AI art. It hurts my heart to see that, but they don't repulse me. Some of them are helping prove the point that AI generated art is not all sunshine and light. I think that's an interesting development.
Let's think for a minute, though.
The AI generators are the proverbial 10,000 monkeys (scaled up far beyond 10,000). After a while one of those monkeys is bound to make some good art. And since the AI training sets are made of other people's art (among other sources), and since so many people are so anxious to help train the monkeys, you're eventually going to get an explosion of uniformly good-looking art. But it must be understood that these AI generators are not here to help artists. They are here to put artists out of business, to obviate artists, to make us obsolete. The most repellent part is that so many people here seem so willing to help that process along.
Deep down, I don't think the AIs will do that. I do think the AIs will take up a lot of the low-level art jobs, and people will still make art. Art is part of our birthright as humans, and AIs don't get to take that away. Worse comes to worst, I unplug my computer and sit there with my airbrush, a paint brush, or even a stumpy pencil (ack!) doing what I still love to do.
There is another side to all this I've not seen mentioned: Star Trek. We've all seen the holodecks on Star Trek and other sci-fi programs. Those things are all powered by AI and follow text or verbal commands. They have databases that likely include the entire cultural and technological output of humanity and maybe some alien races for good measure. They can simulate and replicate pretty much anything. And there seem to be no actual artists in Star Trek. I assume there are some somewhere, but nowhere to be seen in the shows. So, maybe all this AI stuff is a necessary step to get to holodecks?
Food for thought.
You won't see any artists on Star Trek, not because they don't exist, but for the simple reason that we just don't make for very exciting TV. Same with accountants. Oh, they did manage to do something with fashion -- there was that tailor after all -- but would anybody have watched it if there weren't any shenanigans involving alien spies and interstellar wars? We do tend to get overlooked, unless it's a PBS or History Channel special on the Masters, with appropriate scandal and intrigue attached. No, it's all about the adventurers, the scientists and explorers. People wanna see Captain Kirk confront the Klingons, romance the alien princess & save the day. Any side plots involving that guy on Deck 3 who cartoons cute little tribbles all over the ship between shifts end up on the cutting room floor. Gotta make room for the ads, after all.
People will indeed still make art, it's in our basic nature. What changes will be the way we do it. No, it isn't all sunshine and light on the AI side. Neither is it Doom Without End as others are claiming. Surely there's a middle ground, and surely we'll find it. Besides, those 10,000-plus AI monkeys have yet to agree on what hands are supposed to look like in order to produce a reasonable facsimile instead of outrageous mutations, so I think we've still got time.
AI art will likely not replace artists entirely because we almost never can precisely describe the images in our head in a limited number of words. Perhaps if we're allowed hundreds of words in a prompt one day, but a human touch will surely be required to get hand poses, expressions, specific texture details, and such to look as we intended.
I'm not a fan of people making images with AI and then claiming to be an artist - no, you're good at what they're calling "promptcrafting". I'm not even a fan of people using AI art to make money in any way - maybe if a mass amount of concept art is necessary for a large project, but generally, I want artists who personally create their images to be paid instead. For me, as an example, I want to run my 3D renders through AI and reimagine them in different artistic styles. I don't really like the look of 3D, and have always dreamed of being able to make true 2D art as easily as we can make scenes in DAZ Studio. Maybe that could be made possible with a combination of the two methods, I don't know yet. I still wouldn't monitize it because I don't want to take any opportunities away from real 2D artists. In the end, AI art is still a tool.
Most people who know nothing about DAZ products and 3D software assume that the artist created everything in the scene.
Using traditional media.
It's not until they see that everything was made by someone else and just...
Which occurs AFTER you say you did it with 3D models.
I get what you're saying though and it is a good point. I do remember when EVERYTHING 3D-ish was assumed to be POSER-MADE.
Now, in 2022, the assumption is a Game Engine was used.
--------------
So the first assumption is traditional media.
Then it's the digitial simulation/emualation of traditional media.....
Then it's assets and I mean GAME-engine assets....or store bought (since game sites also sell models.....)
----------
The assumption has not dropped to "entering keywords to craft assets" just yet....but there are people who rail against AI and think everything they can't identify as a Daz Asset is AI. lol
So in that respect, it's still the same....but different.
However, what we do as asset purchasers and users (not the actual content creators) is only a few steps above AI image generation when it comes to the work required to create art.
That's the statement I disagree with. World of difference from setting up a full CGI composition and typing in keywords.
The difference between setting up three point lighting and typing in 3-point lighting. Or studying an artist and mimicking their style to typing in their name in the prompt field...
I mean, the ENTIRE AI motif is programming....you insert prompts and you need to learn the keywords.....
@RangerRick "I assume the consumer is using the tools to (in no particular order):"
I said what they do to use the TOOLS, not what they do with the finished work.
-----------------
I remember when the art on Artstation was all about those render engines....everything was cyberpunk and had that same car in that same cyberpunk alley....with that render engine atmospheric fog ( I still wish we had that)...and neon blue-green color. I mean almost every image looked like it was done by the same person. Now, it's all AI-dominated. What a change.
3D as we do it is more akin to set dressing and costuming
without the actors/paid models and physical prop and wardrobe budget
using a modelling app to tweak and kitbash is like the carpenters and other set employees, likewise Marvelous designer etc the tailors, seamstresses
we are the directors and also lighting and photographers
Seriously, are you in a mirror world and how can we get there? ;) Where are you seeing ANY AI art on Artstation? Artstation is where DAZ Studio users go to learn how little we really know about 3D.
Where are you seeing ANY AI art on Artstation?
I'm in the marketplace. One of the many places that 3D asset users go...you know, to get more assets...lol
Artstation is where DAZ Studio users go to learn how little we really know about 3D.
I don't know. You make these broad statements (I thought I was the king of that) about Daz Users and knowing little about...whatever....
Maybe you have some kind of imaginary Daz user in mind. - Who started their 3D awareness with Daz Studio and doesn't know anything else...and just I don't know...is hating on AI from some pretentious pedestal...of being a Daz user....
Maybe I'm reading too far into yout statements.....I could be wrong. But I feel like there's always a zinger in there somewhere.
OK, I have seen a few AI background sets in their Marketplace, but not all that many. Won't argue with you there, as I don't like the idea of selling AI generations either.
All I'm really complaining about here is that people who mainly use purchased content shouldn't really be whining about AI and whether or not it's 'real' art. My Artstation comment was more of a description of a wake-up call; once you really take a look at what they're making over there, you might better understand why many 'professional' 3D artists consider what we do to be so amateurish (but that is a whole other debate that I don't want to start, as I agree with elements of both sides).
No sense either of us arguing here, they're just my thoughts and they're no more important than anyone else's.
I agree, that RTX 40 series great feature is dynamic upscaling of images to the higher resolution using AI,
so the game can run at the lower resolution, but one get high resolution images on the screen.
Will be interesting to see, if this feature (DLSS) make RTX 40 series more desirable, than the previous ones.
Really insightful discussion everyone, I enjoyed reading the various viewpoints.
I really worry for character solutions like Daz Studio if one can make (and you can with dreambooth) consistent characters. An inbuilt style AI in studio would be timely if Daz3D doesn't want to miss the boat of 2d rendered avatars/actors.
Looking at this example, it's beyond what I often get out of Daz Studio and was made in 10 seconds on my comp (after Byzantium model threads from Reddit)
I've actually been browsing ArtStation quite a bit lately in the market to buy some framed prints (retro neon city type stuff) and a lot of what I've been weeding through is AI. Some artists are honest and list Midjourney (et al) in their tools description...some aren't (but it's still obvious). If I'm looking to spend $100+ on a framed print, I'm not looking to buy AI.
The speed at which these systems are proliferating is astounding.
what I think alot of people are missing (or ignoring) is the way non artist “lay people”
will be free to create art for any ephemeral purpose without any need to depend on others.
They do not have to claim to be “Artists’ anymore so than the young people
shooting films at 8K resolution ,on their Iphones are claiming to be “Cinematographers”
This ,I believe ,is the target demographic of this new tech and they will embrace it casually without any of this deep “soul searching" going on in the “artist communities”
That's like the concept of photography and weekend warriors who snap pictures randomly using their camera's auto button and then calling themselves photographers. When I first started getting into photography, I couldn't afford the best camera and the most expensive lenses (and I still can't). But it's not about how much $$$ one can drop on tools. I've seen people throw insane amounts of money into tools and still not produce anything worth viewing all because they didn't care to learn the fundamental basics. That and they didn't have the eye for it.
Just like with DS, a user can simply be that weekend warrior who drops pre-made assets into a scene and uses pre-made poses and pre-made lighting and then hits the render button. And that may be fine for them. Then there are the users who customize everything, right up to and including creating their own morphs and retexturing some (or all) of those pre-made assets. They use their own poses and they use their own lighting. Some may even take their renders into Photoshop (et al) and move on to post work, which again has numerous degrees from simple color/contrast all the way up to overpainting. I've seen pieces created in DS that don't look obviously like 3D and do look like 2D...which is what has always been my personal goal (and in some ways I'm still trying to get there). So there really is a giant gamut of what can come out of DS...and in the end, DS users are definitely not all the same. Far from it.
in other words is NOT your work it's the AI that does the nice job , is this not what I'm understanding ? It would be like me getting you to fix up my stuff then I claim I'm the creator , it's my work, what a great job I've done hmmm...
It is conceptual creation. You make up the concept, the AI executes a visual amalgam of your prompt from what it has "seen" before.
Personally I like the word Cognitive Intelligence, as it's not a self aware/ reflective general intelligence.
Midjourney has gotten extremely popular on my roleplay forums, and more and more I'm starting to wonder why I continue with Daz Studio. I gave up rendering male characters years ago because I couldn't find the content I needed (and I can't just make it, sorry). And now I'm watching Midjourney gurus make amazing male portraits that are completely impossible to make in DS. I'm watching the threads of people waiting for Michael 9, and I can't help but feel that it's just easier and better to use AI for male figures.
At this point, the only advantage I have with using DS is the ability to re-use my dial-spun figures in subsequent projects. Even my personal satisfaction from finishing my creations is diminishing when it gets shoved off the page by a dozen better Midjourney images before the end of the day.
You could still use img2img to input your work through a SD process, having a best of both worlds
Hello @WOLF359 !!!!!
what I think alot of people are missing (or ignoring) is the way non artist “lay people”
I think ignoring is right because no one cares about them. lol
I missed the earlier Star Trek comment about artists......
You won't see any artists on Star Trek, not because they don't exist, but for the simple reason that we just don't make for very exciting TV.
I think the ART shows up everywhere. Whenever they show bedrooms/homes or planets with open areas, there is still art and artsy architecture elements.
Even with replicators, people seem to care about the person who made it...
Even when one of the characters was making an "original" song by mixing different composers to the AI computer, he was still naming people as the input. That's ....fascinating.
I think this answers a fundamental difference between the different logics/goals/aspirations, etc...
Someone wanting to make art because they want the art at the end and someone making art because they want to be an artist at the end.