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
well, i for one, am excited by this and can't wait to play with it.
i rather guess that they sell you each character which the AI composed of a set of morphs. Best they could do is discount it if the product used is in your DAZ library, but i guess chances for that are pretty low...
This sounds more like a text-to-morph generator. Like they're giving ChatGPT a descriptive list of all their morphs (male,female,young,old,etc..), and AI is turning that into a character, via morphs.
My question is, what is the output of what this creates? Is it just a render, a 3D model, etc... What do we get from it?
They're obviously not going to let us download all the resources being used to create it.
A text-to-pose AI generator would probably be more useful for my needs. Let me type in something like "a pose of a running man", and let me download that pose.
I imagine, at the very least, that lot of people currently using AI to generate images might enjoy both the fact that it's not using other people's artwork and the fact that the fingers will always look good.
-- Walt Sterdan
it's really just Smarter online Smart Content which is very dumb and limited due to it's Postgre data base
I have Fooocus ai at home but I choose DAZ. HOwever one can use Ai to upscale or enhance one's images / animations as I am playing with. (Like using Photoshop just quicker and more accurate - no skills needed.)
My issue is Ai generators calling themselves artist which they are not. Yes, we DAZ users uses a lot of models/assets but we change / mold them to create "art". Ai gives you "its" idea of what you the user input into it. Also not every model looks exactly the same in different images it creates. And the biggest one, it just battles to get the fingers/toe's correct.
But the general public, don't care. They think Ai generators are true artists and would pay handsomely for whatever they put out. We humans are becoming lazier and stupid using these tools.
Ai is amazing and scares me at the same time.
I believe this is refering to AI Studio https://www.daz3d.com/ai and is a way of using DAZ product library to generate images.
It doesn't matter. In order for the AIs to work, all the unethical material is already baked in to the AI base. The Daz stuff is just frosting on top. That's how it is with AI because the language and image models require so much material baked in just to do anything. If you're entering text, the AI has to be able to process the text, meaning LLM with all that implies. And the AI is still generating an image with all that implies.
It is what it is. That's all I will say.
Actually I will say one other thing: at this point, if you want to use the ai stuff, you might as well. If you've posted any art or writing online in the last several years, it's a good bet that's been scraped up and dropped in the pot. Your stuff is part of the mix. You might as well get something back for your contribution.
Hey this is a zombie thread! Sorry about that. Teach me to look at post dates.
It's too late now. You've awakened it. Now we have to feed it a virgin or something.
Correction. My company is using ML to build our data model not Generative AI, but I still wonder if AI Studio could not have been built with only the DAZ product library based on the errors I see when I try it out.
https://www.techtarget.com/searchenterpriseai/tip/Generative-AI-vs-machine-learning-How-are-they-different
I don't know what you intend your ML to do or anything about your company, so I can't comment on any of that.
The issue was never the processing method but the source of data to train the model. The whole generative AI boom and likely a fair number of ML projects go back to LAION data models, which were created using unethical data gathering processes and include a lot of material that should not be there. ChatGPT and other LLMs were similarly created using unethical means of gathering data and filled with materials that should not be there. Yet all those materials are there, in the models, and reportedly can't be removed. There's too much money incolved to go back and start over. Furthermore, it's very common in computer programming to build on the works of others, including that work as a starting point. So, to me, it's a safe bet - An underlying assumption, at this point - That anything going on AI or even in ML (if we want to use such granular parsing) is built on data that goes back to LAION data sets or other data that has been scraped from sources without explicit permission. Even now, feeding the data machines is opt-out, not opt-in. See Linked In, DA, and any other large repository of minable data on the internet.
People are amazed by the capabilitiies of the new Deepseek engine, and that may have been trained using output from ChatGPT. The interesting part about that is, Deepseek seems to work just fine. One of the pet theories trotted out of anti-AI circles is that AI models trained with AI-generated data would inexplicably collapse. Yet Deepseek did not collapse. And rhythmically supporting my point above, it represents a big project that was likely built on work that came before it in order to move forward with limited resources.
None of this is an accusation. It's simply an acknowledgement of my final point: Deliberately or not, we all contributed to the AIs we have now - The entire world of text and images on the internet made generative AI possible. Even MLs require enormous bodies of training data. That being the case, we've all contributed and we all have ownership stakes in the resulting AIs. Therefore, if any of us decide to use it, that's perfectly fair. Should there be limitations on that use? That's something the courts around the world will have to figure out, and it may only take the next 5-20 years or so to get an answer.
I have to agree with that!
this was originally about an Ai driven content selection system that hasn't eventuated yet
using actual 3D content chosen through prompting
I too was wondering how this thread got so derailed. Maybe Richard should move all these OT messages to the AI section. The original topic of this thread seem to be pretty dead.