They want to make us pay for characters we've already purchased to use in AI image generation?
Leonides02
Posts: 1,379
[They want to make us pay for characters we've already purchased to use in AI image generation?] I understand Daz is seeing the writing on the wall when it comes to AI, but this isn't the way to go about it and still keep your existing users happy.
Post edited by Richard Haseltine on
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
© 2024 Daz Productions Inc. All Rights Reserved.
Comments
how about if it bacame part of Daz Studio, and allow us to use our existing catalog? To me the sub price would be OK for that. There is no way I would purchase content for this, especially things I may have already bought.
Same. Especially when they end up looking like this...
This is supposed to be Victoria 9? There are zero details.
+1... totally agree with this.
I've done a few 'simple samples'... Daz AI still gets the fingers and eyes wrong. :(
At least her fingers look sort of ok... But the eyes are still funky to me... lol
Agreed. I wonder just how much future profit, and how many long-time loyal users Daz will loose because of this.
Daz should have positioned itself to as an alternative. As people become more interested in telling stories they'll need more control.
Daz Studio is the only way to get that control.
A missed opportunity.
There is a free tier, the fee if for more than 25 images per month and some additional features - remember this is running on Daz' servers (or leased servers, I am not sure) so it costs them money.
I prefer composing my scenes, posing my characters and rendering it out, you know, real effort to produce an image, not press a button and have some randomised image pop out... I want to position the lights and scene, not have them magically appear!
Now, using AI for post work and augmenting your artwork, sure, I'm all for it, but not replacing the joy of creating with automation! I hope DS doesn't become some crappy pay-to-play rubbish, keep them separate, you can have your image generators but keep studio a studio, not some random image generator!
Who would have thought, artists actually like to create art. Agreed, if AI was responsibly and ethically utilized most of us wouldn't bat an eye. But obviously, that's far from the case.
I agree... I must admit i like to dabble with AI... but mostly for texture and fabric shader ideas... It has been inspiring me to create more content for myself. We shall see if this is a good thing or bad for Daz. They are a business after all and they gotta try new things... Only time will tell. I guess they gotta TRY... Ya never know til you try... right?
Indeed, if they used actual artists working WITH the models instead of using artist's work without their consent, then AI would actually be an augmentation rather than a replacement for artists.
The Daz training uses Daz-owned imagery, not PA or member images.
Is it a totally new model? Or, some flavor of Lora for SD15 or SDXL?
- Greg
Where do you see this? Not that I'm arguing, I just don't see that said anywhere.
It doesn't
Ditto.
I like the process of art creation using DS as much as the next person. On the other hand, if I want a quick image using my own content library, I'd like to do that too on occasion. I don't want to have 2 separate libraries of content that are not cross compatible.
Oh, I have no doubt, as Daz...
Well, it's Daz, they have master artists working there, so that would never be an issue, however, I was speaking in general as a whole, but the confirmation is very welcome nonetheless as it's great knowledge to have, as it clears up any question for anyone reading this… ;^)
But in your case you're only using it as a tool, not a replacement for creating your work, so no issue here.
What checkpoint are they using? A LoRA is an addon to a base checkpoint. Whereas a LoRA can be trained on a few dozen images, a checkpoint is trained on vastly more. There have been no checkpoints released that do not have a dubious level of consent when it comes to the use of the art they were trained on. By using a checkpoint that has been developed unethically, any images generated by it are similarly tainted.
And therein lies the rub. How about releasing a loras for Stable Diffusion that we can use on our own computers? Some of us have spent a lot of money building moster render boxes.
https://aituts.com/stable-diffusion-lora/
What I'm trying to understand is where are the clothes, props, environment coming from if you're just using the base (V9 and M9). I've seen other examples which don't even feature the 2 products. The example I've provide shows the inconsisties associated with AI. The T-Rex (not included as a product) wearing her boots while she doesn't have any. The biggest thing is you'll never recreate an AI image to be the exact one you just generated.
There is this in the FAQ under section "How do Products Work"
Our Free and Paid users will have access to our core characters Victoria 9 and Michael 9 with other characters available for purchase from the store in the future that you can add to your account.
I am not sure what that means, but the implication is that you can acquire content to use in the AI engine. It is not clear if you will need to repay for content already owned, or have to purchase another add on licence, like we need to for Interactive, or 3D Printing.
Because someone will share it and the entire point of this will be lost. That's basically what happened to the anime models used by NovelAI, someone leaked them and it led to the public Anything anime model.
Derived AI models are trained based on the main model. If there are flows in this model, they are carried over to the derived model. However, a well-designed derivative model can improve the quality of the main model in some areas. For 3D artists, the DAZ AI model is very useful not only for texture or shader ideas, but also for ideas for posing characters, their relationships, and even entire plots for their future paintings. For example, based on an image of a girl riding a dinosaur in her shoes, an experienced artist would draw such a piece of art on a tablet that buyers would tear it off along with the artist's hands the first time the painting was sold. Really, what kind of devoted pet do you have to be to wear your owner’s shoes? Have you ever had such loyal pets?
They were working with Tafi on a TEXT TO 3D , which seemed promising. I wondered what happened with it.
It seems to me that developing a TEXT TO 3D is very difficult. But if they create it, then the era of 2D AI images will end. Moreover, DAZ developers program in C++. Poor Python wouldn't stand a chance against C++. Of course, this event would have many powerful opponents, since a lot of money is now circulating in AI.