Hi fellow Poser user,

first things first, Marry Christmas, Happy Holidays and a good time to you all!

I recently came across a video on YouTube in which a Daz Studio user created an animation enhanced with AI. I'm not sure if it is plugin of sorts.

I'm curious, is there a user base in Poser who would use AI and should the next version of Poser provide a plugin or interface to connect Poser to an AI generator.

I for myself have decided to ignore AI as a whole. I really don't like the outcome and I think it's a threat to many people's jobs in the IT Business and to Artists alike.

I would love to read what you think about it.

I'm using stable diffusion now.  Ken 1171 has created a plug-in that can export poses to something stable diffusion can use.

Is it a threat? Well, yeah.  This is what happens when the Poser userbase is reduced to those folks that make single image pin up art.

It is easier (and cheaper) to make pin up art in stable diffusion than it is in Poser (or that other program).

The (only) good news is that it is really difficult to do anything long form with SD.  The reality is that there is no "intelligence" in the AI.  Everything is being brute forced.

SD can't do what I do in Poser, so I don't see it as a threat - but if we are being honest since I don't do single still pin up art, I am no longer considered a "typical" Poser user.

Gee, Stable Diffusion is so 2023.  ;-p

Animate Anyone is the new hot sh*t:

https://www.youtube.com/watch?v=8PCn5hLKNu4&t=62s

From a single photograph or drawing directly to a fully fledged animation.

And changing clothing is also no problem:

https://www.youtube.com/watch?v=2_yGx2Mezxg

@ssgbryan

I agree with you on the point you made about the intelligence or the lack there of in AI, but I can't draw the connection between the pin up artist and the rise of AI I use Poser since almost 20 years now, creating erotica most of the time. Since the rise of AI I have seen a lot of pictures done with it and it became clear pretty quickly that this pictures are boring and not very original. They all look the same. I do think AI is dangerous because it is a shortcut to circumvent all the hustling with building the skillset an artist needs to create stuff. The other thing is, in my opinion, that AI was trained by grifting material from the internet without asking for permission or even paying the creators of that material. And it is very much likely that this practice will continue if we use AI more often.

But thanks for sharing your thoughts on this topic. 

@ssgbryan

I agree with you on the point you made about the intelligence or the lack there of in AI, but I can't draw the connection between the pin up artist and the rise of AI I use Poser since almost 20 years now, creating erotica most of the time. Since the rise of AI I have seen a lot of pictures done with it and it became clear pretty quickly that this pictures are boring and not very original. They all look the same. I do think AI is dangerous because it is a shortcut to circumvent all the hustling with building the skillset an artist needs to create stuff. The other thing is, in my opinion, that AI was trained by grifting material from the internet without asking for permission or even paying the creators of that material. And it is very much likely that this practice will continue if we use AI more often.

But thanks for sharing your thoughts on this topic. 

Creating good art with either AI or Poser takes time and effort in learning how to make the best images. There is a sameness in Poser renders as well, due to the fact that you see so many renders that don't go beyond canned poses and "barely there" clothing. 

I'd love to see more realistic skin textures, more realistic characters, more realistic clothing. 

I agree with DeeceyArt.  Unfortunately, the marketing strategy for AI failed to define it as a supplemental tool for a workflow.  It became a lazy man's way of diluting skill sets that take years to develop and master.

I do not think that AI at any time will be affecting the 3D creativities in contrary it will make it even more inspiring making even better models with more and more details. How many times have you looked at these AI creations thinking how great it would be having such posable models in Poser or other 3D software. At the end every artist want to be in control of the creation.

For me after years of extending creation possibilities for poser making bridges, improving the workflow for 3D creations and speeding up the process, using Professional rigging tools and modeling apps, that actually could be compared to AI features getting the most out of 3D models in poser. There is not much difference using a AI engine that can help yo make your models even better.

A while back I Invested into AI local tools, expanding the bridges for model creations in Poser and the results are stunning, not that it is getting faster it is offering more detailed results in the same amount of time. So these AI tools can be used just like a common Photoshop filter or a common Blender Py helping your workflow on 3D modeling and texturing. I am here not talking about Online AI features that might be a little restricted but AI local tools as Internet speed might be placing some limits on the workflow ....

This is a sampler of a work in progress in early stage, a quick bridge shot into poser where Local AI tools in combination with the modeling tools I am using making it possible getting such models in Poser.    

For sure these tools once integrated correctly into 3D modeling will be helping make models beyond any imaginations, not just 2D renders or clips. Putting clothes on a Model like shown in the Clip that Joe Public posted might be the future for the Fashion industry ore online clothes stores where you virtually test clothing before you buy to see if it suits you. This might be rather used in Online clothing stores in the future. But for us artists that make 3D creations it will go a little into another way directly affecting the 3D models in use not interfering to much with the individual creativity .... 

I'm stopping reading that thread, 1000000% tired of all those A.n-I stuff

@Y-Phil

I don't understand, what is A.n-I? Is it just a typo? I doesn't want to polarized with this thread, or make people angry. I just wanted to see how far AI is incorporated in your workflow and if you want to have a AI component build in to Poser.

I'm not a fan of what AI is used for right now. But I get the general idea of it being useful in clothing or hair simulations for example. Thanks again to everyone here involved. So far I take your answers as a yes to AI, but more as additional tool, not as a substitute for Poser. 

I hope we can get some more voices. Those images above are very interesting. If it doesn't get in conflict with work in progress secrets, it would be helpful to know what part of the workflow is made with AI and how do you decide if and how you get AI involved.

I can see AI being used to beef up the older cloth, hair, and face room stuff. I almost wonder if AI can be used for what used to be called "universal poses." The only caveat with that is there are so many different bone hierarchies used in Poser figures (which was part of the reason "Universal Poses" didn't always work.)

Not to post a render from "the other software" to create an argument, but the new Face Transfer 2 for DS does use AI in part for generating the morph and texture. It does a pretty decent job. Here's an example of G9 using Face Transfer and some additional face morph tweaking. I think "key" here, though, is starting off with a base figure that is more generic than stylized. It only works with G9, which is basically a blank slate for both male and female characters.

I'd love to see Poser moving forward again to recapture and surpass the magic it had in its prime.

The advantage atm is to be able to make a render 10'000 x 10'000 without any quality loss a great feature for printing, AI generates the appropriate textures automatically enhancing wanted details for the needs. Sending through a Macro the textures into AI returning and overwrite the previous. this is practical avoiding having to reassign the textures to the model. You need to find a balance as such a texture can reach 150 meg up really fast. Further is to enhance the detail level of a mesh with AI brushing , still working on this ...

So AI in this sample is changing the way of skinning giving a all new experience in your renders, Not talking about Postworking your Image but how the skins are and how they will render

Here is a sample of what happens if I run the texturing through a AI filter command raising some of the details and how Poser will render it increasing the maps to 8k. It naturally depends on your taste how it will result. More realistic, more like a painting, or a comic style, this also can speed up your workflow not having any further needs of  filters or time consuming Postworking on your renders

I think Poser 14 has certain AI developer options in early 2024. Such as...

* Renderosity buys out Ken's Openpose plugin and then ships it as standard for free. This gives a market advantage, especially if it works with a $50 Poser 13 (I assume another 'perma-deal', like the current $50 Poser 12 offer). DAZ Studio only has a 'no-hands, no face' Openpose option. Thus Poser would have a market advantage, especially if the Openpose plugin is allowed to work with a $50 Poser 13. Poser devs don't have to maintain anything there, as the plugin just works.

* Allow PostFX to have any Python script plugged into it, for further post-render processing. This would allow AI connections, yet AI-phobic people wouldn't have to use it or see it. Again the Poser devs don't have to maintain anything, just open up PostFX to third-party scripts.

* 'Text prompts to AI-made strand hairstyles for 3D' is coming soon, but it's early days - that is probably something for Poser 15.

Anyway, here's a nicer AI picture than the ones above. Stable Diffusion 2.1, from a text prompt without any image-to-image. Took about ten seconds to render. But with Poser I can set up basic real-time renders, and then next time use these as guides... to get a similar picture composition that's exactly right.

@ HartyBart

These Online apps can be quiet interesting but rather limited on the output, I mean size .... try scaling up your Image for a Print and get it like this.

You might want to open the Image into a new window to see the full size if you can :) 

I think this is the Point where AI can be directly Helpful in 3D applications making your renders useful and your Models. You can enhance any " old " model making it top notch right in Poser rendering with fully enhanced details, reducing render times from hours to seconds. The Image above has been re rendered in 5 seconds from  950x534 to 11400x6408 and I mean re rendered 

I actually do everything Locally, do not like having to work online but costs a little  more, needs a little more space, uses up quiet some time setting up the programming for own benefits. So I would rather think not something to be Integrated into poser it self it would not be possible for the crew giving support for such engines, Linking these or making bridges is the best solution. But this can't be made with just a plugin as some would like it some others would prefer something else for there needs ...

This might be something for 3D artists helping to improve there works if they want it to look realistic. Not really my style so I'd rather skip that one but I believe that allot might find it a interesting tool for there works without affecting there creativity  .....

The Most INSANE AI Image Upscaler

so actually https://magnific.ai would sure do a good job for these realistic results on your final renders I guess it would be more for the professional needs and still it remains your render your work as you just make a enhancement or an enlargement of  your art. 

Lets go a stepp further taking the sample of " magnific AI " witch could be a really useful tool for professionals or for Poser Creators. Not using this tool for enhancing the final Image but using it to enhance your doll textures giving them incredible realistic details. Your render will not need any further fix, your doll will not look like a silicone doll any longer in poser after rendering. This might be the Part for creators that know how to make the textures and know, the more detail they put on a skin map the better it will be on the doll, it also would give the opportunity not to be limited in size, easy to make 8k skins. Need of permission to use a real skin from a Photo " Nop " Need a real model for photo shooting " Nop " you create your detailed skins your self and you are the owner.

you tell me how would La Femme look like using this methods to make textures ?! and sure the creators would be having more fun with stunning results. this is the part where the artist needs to decide if he wants to be using these tools to make better things, not the task of Poser or the creators of Poser ! Tools to make better figures are given so it is up to the creator to use these, if they want to have improvements in there Products.   

I have been playing with Poser for over twenty years and came to it late in my life so I guess I am in the 'you cannot teach old dogs new tricks' point in my life.  For me the discussion of AI in Poser is really just a reincarnation of the 'make art button' that so many have been seeking for years.  I personally have no interest in it with regards to Poser, although I am very keen to improve my skill set in Poser as I suspect I have only scratched the surface.  One aspect of AI I have a problem with is that many pictures are created by using art work already created by someone else and that really does not lay easy with me.

I have been playing with Poser for over twenty years and came to it late in my life so I guess I am in the 'you cannot teach old dogs new tricks' point in my life.  For me the discussion of AI in Poser is really just a reincarnation of the 'make art button' that so many have been seeking for years.  I personally have no interest in it with regards to Poser, although I am very keen to improve my skill set in Poser as I suspect I have only scratched the surface.  One aspect of AI I have a problem with is that many pictures are created by using art work already created by someone else and that really does not lay easy with me.

AI could be a useful and valuable addition if it were trained for prime ingredients such as textures and poses/movements between poses.

I'm trying to figure out exactly how AI texture enhancement would work. Typically, you'd have (at least) diffuse, roughness, and normal for each texture map. Multiply that by, say, 5 material tiles (like in La Femme) and you have 15 textures to recalculate. And that is just for one character.

Once that is done, then you have the scene render, which would have to calculate all the light bounces for each item in the scene. And the number of bounces would have to be calculated for each light in the scene. And it wouldn't make sense to do a 4K render for a scene where you enhanced the textures by four times the amount, so render times would be increased, yes?

So, wouldn't that end up being much slower a process than just having AI do a post-process on a finished render to make it larger?

Saying that AI would be a useful tool for creators brings up the obvious.

You can not claim anything you created with AI is yours, or sell it, give it away, etc. 

So this is a very slippery slope. So slippery in fact, that some obviously will use and do just that. And are justifying doing exactly that. Think about that for a bit.....

It is no different than taking a game ripped mesh and textures, modifying them, and rigging it for Poser and selling it as your own creation.... It doesn't matter what you end up with after when doing so, it is still based on an illegal derivative. Making it illegal, and stolen work from the start......

Opinions don't equal legal, they never have either. Using generative AI to make content is not legal, in multiple ways.

This thread is slowly turning into a perfect "Here is how you could get away with it", "And who cares if it isn't legal...." example...

Which in itself, violates the TOS...............

Tread carefully on this one....... 

DeeceyArt posted at 11:16 AM Wed, 27 December 2023 - #4479652

I'm trying to figure out exactly how AI texture enhancement would work. Typically, you'd have (at least) diffuse, roughness, and normal for each texture map. Multiply that by, say, 5 material tiles (like in La Femme) and you have 15 textures to recalculate. And that is just for one character.

Once that is done, then you have the scene render, which would have to calculate all the light bounces for each item in the scene. And the number of bounces would have to be calculated for each light in the scene. And it wouldn't make sense to do a 4K render for a scene where you enhanced the textures by four times the amount, so render times would be increased, yes?

So, wouldn't that end up being much slower a process than just having AI do a post-process on a finished render to make it larger?

you do not have to run Bump maps through the AI that is not a good Idea normally you just do it on the Diffuse map, maps that affect the high detail of your render. Getting a good diffuse map with loads of thin details will also solve allot of problems with the bump map that might not even be necessary any longer, the skin does not need to be bumpy and look like it been hit by micro meteorites, Mostly an issue on releases that the artist is trying to add details with bump because he's diffuse map is just flat with only minor skin details.

Sure you can run a post process on the render enlarging the details this would be the individual choice. But for a creator that sells a product it might be better if he is able to offer a better texture if possible, would not be a good Idea if the creator would say, Hey guys I offer the basic skin for 16-20$ but if you what a good skin get your self a AI account on magnific for a 200$ .

Customers expect a good quality adapted to the time else they will just jump onto something else, poser just cant offer textures that are built like a decade ago with the new optix features, this would be a no go and sure not very productive. 

I do not say it is a must ... Nop ... it is just a hint, I could not teach you any better, artists are very Individual. Some will take the opportunity and be ahead with there learning curve. If just one starts offering better skins by using advanced tools, what do you think will happen with the creators that don't ?

I started a month ago installing the process on my computer with all new filters, this is just the beginning, I already done it that way since decades with bridges and other type of rigging. If I can do better then I will, if I can I'll get the tools to be able to offer the best I can, this keeps my customers happy and they keep on supporting me, my sites and Poser.

Again .... not a must, just some hints, Just sharing my own experiences as a creator that might help a little others. A litte success is like the fuel for an engine, it keeps motivating the artist to continue. But the artist needs to keep up with time to earn that fuel.  

I was thinking more on an actual 'intelligence' side: on a garment recognise the areas where wear and fading occur and given the data of the fabric apply to the 'as woven' seamless textures the effects thereof. 

shvrdavid posted at 12:46 PM Wed, 27 December 2023 - #4479653

Saying that AI would be a useful tool for creators brings up the obvious.

You can not claim anything you created with AI is yours, or sell it, give it away, etc. 

So this is a very slippery slope. So slippery in fact, that some obviously will use and do just that. And are justifying doing exactly that. Think about that for a bit.....

It is no different than taking a game ripped mesh and textures, modifying them, and rigging it for Poser and selling it as your own creation.... It doesn't matter what you end up with after when doing so, it is still based on an illegal derivative. Making it illegal, and stolen work from the start......

Opinions don't equal legal, they never have either. Using generative AI to make content is not legal, in multiple ways.

This thread is slowly turning into a perfect "Here is how you could get away with it", "And who cares if it isn't legal...." example...

Which in itself, violates the TOS...............

Tread carefully on this one....... 

As long as you use AI like common PS filters you are on the secure side. If it were your way then the whole Internet is Violating the TOS ..... Look around you, they would have to shot down any Artist sites on the Internet for Investigations and now even worse just because of that Illegal AI !!! Wonder when the FBI will put an arrest order to Microsoft or Google !   

Wonder when the FBI will put an arrest order to Microsoft or Google !   

FVerbaas posted at 1:04 PM Wed, 27 December 2023 - #4479656

I was thinking more on an actual 'intelligence' side: on a garment recognise the areas where wear and fading occur and given the data of the fabric apply to the 'as woven' seamless textures the effects thereof. 

In the latter, all of the detail is sculpted into the high res mesh and then the normal, cavity, and roughness maps are generated by the differences. It's the cavity maps that control where the wear and tear appears, but the cavity maps typically aren't distributed, they are only used internally by Substance Painter. The diffuse, roughness, and normal maps are used to create the final fabric, which will respond correctly to light positions.

@JustBeCause

you said:

you do not have to run Bump maps through the AI that is not a good Idea normally you just do it on the Diffuse map, maps that affect the high detail of your render. Getting a good diffuse map with loads of thin details will also solve allot of problems with the bump map that might not even be necessary any longer, the skin does not need to be bumpy and look like it been hit by micro meteorites, Mostly an issue on releases that the artist is trying to add details with bump because he's diffuse map is just flat with only minor skin details.

Baked in highlights and shadows will not respond correctly to lighting. For that you either need sculpted in detail (which more than likely will need subdivision), or really good bump/displacement/normal maps. Depending on diffuse alone will not be adequate for a physically-based render engine.

"Hey guys I offer the basic skin for 16-20$ but if you what a good skin get your self a AI account on magnific for a 200$ ."

How many typical Poser customers do you think will be willing to pay $200 for a character texture?

DeeceyArt posted at 2:43 PM Wed, 27 December 2023 - #4479660

Ah yeah, I see. When generating texture sets in Substance Painter, the wear and tear effects are generated from the normal maps, or by comparing a high res mesh to a low res mesh.

In the latter, all of the detail is sculpted into the high res mesh and then the normal, cavity, and roughness maps are generated by the differences. It's the cavity maps that control where the wear and tear appears, but the cavity maps typically aren't distributed, they are only used internally by Substance Painter. The diffuse, roughness, and normal maps are used to create the final fabric, which will respond correctly to light positions.

FVerbaas posted at 4:11 PM Wed, 27 December 2023 - #4479663

DeeceyArt posted at 2:43 PM Wed, 27 December 2023 - #4479660

Ah yeah, I see. When generating texture sets in Substance Painter, the wear and tear effects are generated from the normal maps, or by comparing a high res mesh to a low res mesh.

In the latter, all of the detail is sculpted into the high res mesh and then the normal, cavity, and roughness maps are generated by the differences. It's the cavity maps that control where the wear and tear appears, but the cavity maps typically aren't distributed, they are only used internally by Substance Painter. The diffuse, roughness, and normal maps are used to create the final fabric, which will respond correctly to light positions.

Very clever but it sounds like the modeler still has to model the effects, and some clever but non-intelligent procedure is applied to fake the texture based on the mesh. I was thinking AI to do one or two steps beyond that: 'I recognise this as a geometry representing a coat. Typical wear on an old coat is the collar and the cuffs. On the collar combined with grease from the hair, on the cuffs combined with whatever came about, so that's what I apply.' 

As a matter of fact, I think you can create cavity and curvature maps in ZBrush too. Maybe you can take it into Poser and add some sort of wear and tear layer on a material, using the cavity map to drive where the worn areas appear. Never tried it, might work.

See this link for generating cavity and curvature maps in ZBrush.

After reading that, it might be curvature that you'd use for the wear and tear detail.

DeeceyArt posted at 2:47 PM Wed, 27 December 2023 - #4479661

@JustBeCause

you said:

you do not have to run Bump maps through the AI that is not a good Idea normally you just do it on the Diffuse map, maps that affect the high detail of your render. Getting a good diffuse map with loads of thin details will also solve allot of problems with the bump map that might not even be necessary any longer, the skin does not need to be bumpy and look like it been hit by micro meteorites, Mostly an issue on releases that the artist is trying to add details with bump because he's diffuse map is just flat with only minor skin details.

Baked in highlights and shadows will not respond correctly to lighting. For that you either need sculpted in detail (which more than likely will need subdivision), or really good bump/displacement/normal maps. Depending on diffuse alone will not be adequate for a physically-based render engine.

Personally I do not use the AI Online features I run it locally with AI models that are local on my machine, no need to grab references or models from the Net. Online AI engines are rather built to generate changes on your Image with words that cause a Algorithmen adding details from other references to your Image up to completely changing the original, this also might not be very helpful for making 3D maps :) 

Well that is the point Poser/DS customers will sure not be willing to spend 200$ for limited credits using a AI to upscale there renders , unless it is for professional use. So there comes the Creator that if possible finds good solutions so that there is no need of upscaling the product he is selling, offering a product with satisfying details for the end-user.  

JustBeCause posted at 1:18 PM Wed, 27 December 2023 - #4479657

Wonder when the FBI will put an arrest order to Microsoft or Google !   

You are confusing creating the tool, versus the legalities of using it. 

Besides:

There is not a big difference of using a 3rd Party model for a render if you have not completely made it your self or using a AI for a render that is using models to generate it. for both you pay to use the models, at the end you do not own them unless you made the models your self, even if using a 3rd Poser DS model to morph or texture it, you just made a derived art on a base, the original creator keeps all rights. In theory you have more rights on a AI made product then on a 3D render in witch you do not own any of the used models. But in both cases AI or a Poser/DS  render with models created by 3rd party you will not be able to put a full copyright on the renders. if it is a comic you can place a copyright on your text, or your signature is copyrighted but the render unless you have created it all by your self, is almost no chance to registrate it under your copyright ownership. There is somewhere a text that is saying that DAZ is legitimate owner of any Derivative creation made with Millenniums or Genesis, so even there it is impossible Placing a full copyright ownership on any creations based on these figures!      

Concluding because both way's  are unique derivative by a minimum of 5% it is a legal creation and not theft, but you are not the copyright holder and will never be. Not possible to make a registration of full ownership.

"you are right , at this point you need to find a balance a AI upscale can't handle a bump map correctly, it is what I meant with not a good Idea best is just to leave your bump as it is or make a larger one with a converter also depending if you want to use a normal map or a bump map Photoshop is having these features so instead of a AI you just use a 3D filter. "

Unfortunately, when I try to use the old 3D filters, I get a message in Photoshop 2023 that they are being discontinued. Not sure what will appear in their place. Although, since Adobe now owns Substance Painter, which can generate all those maps from diffuse texture, they will probably just rely on that rather than including them in Photoshop.

JustBeCause posted at 4:51 PM Wed, 27 December 2023 - #4479668

shvrdavid posted at 1:33 PM Wed, 27 December 2023 - #4479658

JustBeCause posted at 1:18 PM Wed, 27 December 2023 - #4479657

Wonder when the FBI will put an arrest order to Microsoft or Google !   

You are confusing creating the tool, versus the legalities of using it. 

Besides:

There is not a big difference of using a 3rd Party model for a render if you have not completely made it your self or using a AI for a render that is using models to generate it. for both you pay to use the models, at the end you do not own them unless you made the models your self, even if using a 3rd Poser DS model to morph or texture it, you just made a derived art on a base, the original creator keeps all rights. In theory you have more rights on a AI made product then on a 3D render in witch you do not own any of the used models. But in both cases AI or a Poser/DS  render with models created by 3rd party you will not be able to put a full copyright on the renders. if it is a comic you can place a copyright on your text, or your signature is copyrighted but the render unless you have created it all by your self, is almost no chance to registrate it under your copyright ownership. There is somewhere a text that is saying that DAZ is legitimate owner of any Derivative creation made with Millenniums or Genesis, so even there it is impossible Placing a full copyright ownership on any creations based on these figures!      

Concluding because both way's  are unique derivative by a minimum of 5% it is a legal creation and not theft, but you are not the copyright holder and will never be. Not possible to make a registration of full ownership.

If you are using AI in a stand alone mode I can see the use but to let an AI engine trawl the internet for images made by others and then use them, without the owners permission, to create a final image based on a few words is cheating/theft.  If someone was to download couple of e-books and used AI to merge the two books into one you are open to prosecution if it is published as the person never used their imagination to create the story.  I see now difference if it is an image to text.

Actually a big game company that is working on serious base will sure buy a Model but they will by the full ownership, avoiding later issues. If you buy a Licence to make a render then you are allowed to use the model in the render but you do not own the model, causing that you can't claim full rights on the Image you produced, you might own the story but not the character, at this point it is difficult having your Image registrated under full ownership in a copyright office. One might ask, who is that character, your response, Victoria, question, whom belongs Victoria? Sure the render is yours, your collage,  but you are not the intellectual owner of the whole. The more premade stuff you use the less it is that you are the legitime creator. Unless the creator gives unrestricted permissions of usage on derivative works and usage of the Models. but sure is not the case with stuff you buy for Poser DS.

If you go hardcore, you use La Femme make a render with her and name her "Aisha" is it then Aisha or is it La Femme you used ? So no matter how many Morph Packages you will create to sell, no matter how you will name these it will always be la femme and the Intellectual owner of this character will always be the creator, no matter how and for what you will use it. It will just be a derivative of the base in witch you only own a basic Licence of usage. 

A simple question ? What is it that you really own on your Computer ? Do you actually own any part of it ? I guess not, sometimes I even wonder if we own our own flesh or if someone is calming rights on it :)

Just Imagine as a extreme ... you produce a render with La Femme, a doll package, a Pose package. you make a Print, on that Print you claim full ownership and sell it for a million on a auction .... I bet that at this point some will start to sue you :) as your render license in most cases is just for Personal usage and so is your render and Print, you just are not the Intellectual owner unless you created it all by your self or purchased full owner and usage rights.

JustBeCause posted at 9:15 AM Thu, 28 December 2023 - #4479688

Actually a big game company that is working on serious base will sure buy a Model but they will by the full ownership, avoiding later issues. If you buy a Licence to make a render then you are allowed to use the model in the render but you do not own the model, causing that you can't claim full rights on the Image you produced, you might own the story but not the character, at this point it is difficult having your Image registrated under full ownership in a copyright office. One might ask, who is that character, your response, Victoria, question, whom belongs Victoria? Sure the render is yours, your collage,  but you are not the intellectual owner of the whole. The more premade stuff you use the less it is that you are the legitime creator. Unless the creator gives unrestricted permissions of usage on derivative works and usage of the Models. but sure is not the case with stuff you buy for Poser DS.

If you go hardcore, you use La Femme make a render with her and name her "Aisha" is it then Aisha or is it La Femme you used ? So no matter how many Morph Packages you will create to sell, no matter how you will name these it will always be la femme and the Intellectual owner of this character will always be the creator, no matter how and for what you will use it. It will just be a derivative of the base in witch you only own a basic Licence of usage. 

A simple question ? What is it that you really own on your Computer ? Do you actually own any part of it ? I guess not, sometimes I even wonder if we own our own flesh or if someone is calming rights on it :)

  

To me the difference is clear, I gather together all the different components to produce, in my case, an illustrations for a book.   I can use that illustration in my book without asking anyone's permission because I already own the image.  To create that image can take me days, sometimes weeks but the creation of the image is my work.  On the there hand if I use an AI engine and describe a scene which it builds for me by taking snippets from many other images that I had no hand in creating I cannot say with any honesty that the image is mine.

I am sure there are many that will try and muddy the water in relation to AI but the fact remains others people's work is being used without their express permission, which I find morally wrong.  That said I do accept I come from a era where most people had a strong moral compass. 

I sure do understand the point about renders using 3D assets, but also have to try giving a understanding for the AI that is coming in really fast, not limiting my self there as for sure they are going to adapt all the rights else economy might stand still. Serious AI programs are learning algorithm's not like most think, it grabs Images on the world wide web rips them apart and puts them back together, this is simple mind thinking to be able giving a explanation on how it might be working. 

A good AI collects Information then it builds up a so called model, you can compare it with a mathematical function of your brain, but they call that file a Model witch is one of many learning algorithm's comparable to a .Json file Not a Image it self, the learning process with given Information generates a sort of Intelligent Filter, just like you use in Photoshop just that it is a learning filter. You can also compare it with a encyclopedia of Information.

So In theory human kind is creating a big brain, not local like most might imagine, there is no shut down trigger on one place it is working like a assimilations of all computers that are online collecting Information, the integration into your system has already been done, there is no turning back on this one, the AI models are often generated automatically on your system that will be to send and receive the appropriate Information's. Actually most digital devices are getting integrated.

The only thing that needs to be adapted is the rights but this for sure is going to be solved quickly as none of these features have been locked until now, so what remains is either one starts learning to live with it, learn how to use it, or shut down any Digital device.

But to make it understandable it is not a rip of from renders or pictures throughout the Web, they are individual models created with Information's that the Users are giving, it is how the AI is learning . 

Ok it is a little like the Terminator Movie, the beginning, where mankind is trying to play god and will loose control in one way or another scary it is but nothing we really can change about this. It is still up to the individual to decide weather to use it or not atm. Later it will be a full integration of your digital tool there is no way around it and none will ask anymore where the AI has gotten the info to make the models as this for sure can not be controlled by Humans.

So this is the simple explanation about the basic function, Immoral or not it is already a part of the whole, but basically this long text is just to explain that the generated Image is not a rip of an Image, but it is a model generated from Information's that the users have given. Same like a Homan that grows and learns with Information he is receiving to accomplish a task.  That is why it is called AI, non Humanoid learning being.  

JustBeCause posted at 11:35 AM Thu, 28 December 2023 - #4479696

I sure do understand the point about renders using 3D assets, but also have to try giving a understanding for the AI that is coming in really fast, not limiting my self there as for sure they are going to adapt all the rights else economy might stand still. Serious AI programs are learning algorithm's not like most think, it grabs Images on the world wide web rips them apart and puts them back together, this is simple mind thinking to be able giving a explanation on how it might be working. 

A good AI collects Information then it builds up a so called model, you can compare it with a mathematical function of your brain, but they call that file a Model witch is one of many learning algorithm's comparable to a .Json file Not a Image it self, the learning process with given Information generates a sort of Intelligent Filter, just like you use in Photoshop just that it is a learning filter. You can also compare it with a encyclopedia of Information.

So In theory human kind is creating a big brain, not local like most might imagine, there is no shut down trigger on one place it is working like a assimilations of all computers that are online collecting Information, the integration into your system has already been done, there is no turning back on this one, the AI models are often generated automatically on your system that will be to send and receive the appropriate Information's. Actually most digital devices are getting integrated.

The only thing that needs to be adapted is the rights but this for sure is going to be solved quickly as none of these features have been locked until now, so what remains is either one starts learning to live with it, learn how to use it, or shut down any Digital device.

But to make it understandable it is not a rip of from renders or pictures throughout the Web, they are individual models created with Information's that the Users are giving, it is how the AI is learning . 

Ok it is a little like the Terminator Movie, the beginning, where mankind is trying to play god and will loose control in one way or another scary it is but nothing we really can change about this. It is still up to the individual to decide weather to use it or not atm. Later it will be a full integration of your digital tool there is no way around it and none will ask anymore where the AI has gotten the info to make the models as this for sure can not be controlled by Humans.

So this is the simple explanation about the basic function, Immoral or not it is already a part of the whole, but basically this long text is just to explain that the generated Image is not a rip of an Image, but it is a model generated from Information's that the users have given. Same like a Homan that grows and learns with Information he is receiving to accomplish a task.  That is why it is called AI, non Humanoid learning being.  

Many predictions of the future are based from a very biased position such as the common belief that some have that everyone lives their life on social media and the life revolves around the mobile phone.  Other than this and one other art forum I am not involved in social media.  I am of an age where I do have a smart phone that I switch on to make a call and then switch off again and because of that many of the predictions of the future are not what I see life becoming for me.   AI is just the latest craze that will create problems for the legal profession and will take years to resolve, if at all, in much the same way the profession is struggling with the impact of social media.

Happily I am of an age where little of this will touch me and I will be able to continue living while taking full account of my own personal moral compass.  My only concern is those that  love that will have to live with the world as it is but i did my bit by not having children. 

I have watched and held my breath. I only speak up because there is a lot of misinformation here. JBC you have no idea of which you speak. You have no idea how a diffusion model is produced or deployed. The waters get muddied by disinformation. Portraying yourself as knowledgeable in this field is laughable at best and quite a dis-service to those of us who are data scientists and imaginative artists.

"A good AI collects Information then it builds up a so called model, you can compare it with a mathematical function of your brain, but they call that file a Model witch is one of many learning algorithm's comparable to a .Json file Not a Image it self, the learning process with given Information generates a sort of Intelligent Filter, just like you use in Photoshop just that it is a learning filter. You can also compare it with a encyclopedia of Information."

What kind of BS is this? Really? Would you really like to know how a diffusion model is built? Not your way. An "A.I." doesn't "collect information" ;and exactly how you equate this to "many learning algorithm's (not a possessive btw) comparable to a .Json file." makes no sense what so ever.

We've had this conversation before. More than once. It has been shut down for various reasons each time. Usually the misinformation and fear of the future arguments win the day. I understand that. There are unanswered in the court of law arguments currently underway. They will fall on both sides of the fight as there is more than one way to build a diffusion argument and many companies are building their own using GANs and not the addition/subtraction of noise method currently in use by Stable Diffusion and others.

So stop with the BS and try reading a white paper or two. I have linked to them in the past and got in sh!t over it so I will not do so again. There are reasons that Ken and others have stepped out of these conversations and folk with knowledge like myself fear to tread.

@parkdalegardener

Guess this would be your language explaining a Diffusion  " Model " in short :)

" Diffusion models work by training an artificial neural network on images and their text descriptions. These are computer systems vaguely inspired by biological neural networks "

But most people would not be able understanding the signification of this text ! same like placing a mathematical Formula to explain the algorithms

Again; nope. Vague crap definition by whom? Somebody on TicTok? Wasn't me so don't try and put words in my mouth.

"But most people would not be able understanding the signification of this text ! same like placing a mathematical Formula to explain the algorithms"

Huh? the signification of this text. Really? Reminds me of the Homer Simpson quote. "smrt"

Diffusion models are built by repeatedly adding noise to an image until the training image is beyond recognition. The reverse generates an image.

Example is a photo of a cat. The cat image has a predetermined amount of random noise added and the label ( the text part of the model) says it a cat. There is more noise added and the label is cat. More noise is added and the image is labeled cat. Soon there is only static noise in the training image and it is labeled cat. That is your "model". A set of rules for denoising an image. To get an image you type c a t and a random noise is generated and the denoising starts. The operator (you) determines the strength of the denoising logarithm and the number of iterations it makes in an attempt to create the request. The result, for better or worse; is "cat"

That's it in a simplified nutshell as diffusion models now stand. There is a massive legal mess happening around this method as we have all heard. Well the method is changing. GANs are being used to train now. General Adversarial Networks. This is closer to autonomous driving training than current diffusion models, but that is beside the point and a different conversation yet again.

I'm not responding any more to this. The subject is drifting off topic into personal attack which is against the TOS.

So my summary would be, for the moment at least, if you don't know what you are playing with leave it alone and, if you really know your stuff, tread with care.

Topaz or HitPaw for instance,

are good AI tools that one can use in combination with poser these also work locally with models that get downloaded after install, here you do not need to worry. They give a good result scaling up your renders or if you want to enhance your older textures. At this point no need to go into deep technical matter and you can work offline. Might even remove the need having a poser integration! there are also some other programs but I think these have lifetime licenses ....

Read the terms at the links posted above. And better yet, understand them.

If someone uploads someone else's texture that is not theirs to any site to upscale, they more than likely just gave someone else's work to that company based on the terms.

So actually, you do need to worry about it. If it is not yours to give away the rights too, don't give it away as if it is yours...

If you do it on your system with a local model, and then the model/program uploads it to their server, the same applies. And many models, do just that, by design.....

They can't make a dime giving away models they can not profit from.....

People need to understand the terms of using these tools. And it is very obvious that some people don't have a freaking clue what the terms they agreed to actually mean.

These are quotes from the above linked Terms of Service, typo's and all....

"And you hereby grant us a worldwide, non-exclusive, non-sublicensable, royalty-free licence to reproduce, duplicate and display your artwork created in the course of using our website/products/services for the purpose of displaying our website/products/services or for commercial promotion."

Pay close attention to the non sublicenseable part..... Better yet, understand what that factually means in legal terms..... It is no longer yours to license, sell, distribute, etc, ever again..... Even if it was yours to begin with.......

"You shall not assign any of your rights or obligations under the Terms in any way without our written consent. To the extent permitted by law, we may assign our rights and obligations without your consent."

What do you think the chances of getting your rights back actually are after you ask for written consent? Try it and see......... Again, they are out to profit from you using it, not the other way around...... Best of luck with that......

And finally, this part.... Which basically means they can say anything you created with it, is theirs, after the fact, based on the stuff above...........

"We reserve the right, in our sole discretion, to make changes or modifications to these Terms of Use."

These are legal terms, and they are not opinions on "here you do not need to worry"

Because that is not the case at all...............................................................................

@ shvrdavid

The applications above are local you have no access to the world wide web to get Images they do not have access to yours, very simple, a tool like a filter in photoshop. you can choose either grant access to your works for there learning purpose, or not . You only work with your Private Renders Images and you sure do not lose your rights. else it would be that even if you use a photoshop filter you would loose any rights on your works! These tools have direct filter plugins in Photoshop, no way for them to access your works, you do not upload anything :)

Btw Making works based on a 3rd Party doll you loose all the rights on your works  :)  Go once through the guidelines of Daz ....  I wonder if La femme has a guideline that allows making Derivative work on and with her, if it is written that you are allowed to sell derivative works from such a doll ?! Or is it so that it is written that it is not allowed ?! Think about that one.

I guess that all this In here and at Daz are the main purposes why Creators slowly are running away looking for better places for there Art.

You seem to be missing the point, you already agreed that what you make with it is theirs when you downloaded and used the software...

It is in the Eula you agreed to. And your opinion on that doesn't change the legalities in any way.

Jup I Read the guidelines, the ones you have posted are the ones of Usage of there site :) there is a big difference. I cant Imagine that Disney, Lucas Warner etc would be using it if they could take there rights :) Or you just had a Look at the Android app and online Image creation guidelines. but not the Image enlargement filters. It does not matter, You can use the old fashion way others will Improve. Fearing the new does not bring you any further, No Improvement's no sales that is the rule.

All you say gives me rather the Impression of a fear not being able keeping up, at the end you for sure will have to find a good solution if you want to redistribute your works or get a little attention on them, no way around making your works way better. Others will that's for sure :)    

Jup I Read the guidelines, the ones you have posted are the ones of Usage of there site :) there is a big difference. 

I look at the legal side of it. I was referring to the ones you linked and then stated "you need not worry"... The same Terms of Service you now claim only apply to the site, and that is not the case either.... 

Again, I will quote from the Terms of service you obviously don't understand.....

"You are reading the terms of service ( the "Terms"), which govern the relationship between users ( hereinafter referred to as "you", "user", "Licensee") and HITPAW CO.,LIMITED (hereinafter referred to as "Hitpaw","Licensor", "we", "us" or "the company") and set forth the terms and conditions by which you may access and use our related websites/PRODUCTS/services. It shall have the same legal force for both sides."

Note the product part, which includes the programs you download from the site, and anything you make with them falls under it..... 

You stating that it only applies to the site, only proves my point further. It applies to everything on the site, including the local model(s) you can download..............

Your legal advise is not so good. Truthfully, it is terribly misguided....

Maybe it is a language barrier, dunno. But your interruption is very wrong......

Well that is ok Hitpaw is not a part of my system (" You just seem to Pull out extreme samples to block any possible Improvements ") these are samples on how one could Improve there works without having the need to wait until Poser decides integrating Features that might help enhancing Textures and renders, witch I doubt it ever will happen. So actually the Idea was trying to help users with experiences I am having. Not that I am Intending showing my actual workflow on figure creations, even that it could of been quiet helpful for a Poser community witch is slowly fading away. 

I'v seen many in here in the past that were trying to share a little there workflows and how one could do Poser works in a more productive way rather then having to many efforts on there creations, Fact is that it ends with big rejections and fears having to go away from there habits that are almost 2 decades old.

Not even the new features from Poser 2014 have ever been used because of these problems so creators end up keeping these as there little secrets." So see I just Placed two samples or three I think in total ", Not reviling my actual built to see the reactions, and they are how I was expecting them, was hoping a little that this poser community might of changed a little having better ears for the new since Poser is also getting Newer, but sadly this has not happen and probably never will. 

Sometimes I learn In here some features that are quiet helpful for a better workflow but these are quiet rare as the ones who could give really good advice have also experienced these issues, tracking off as these do not really deserve help of Improvements in figure creations.

The mentality in here: If a Painter is painting a Portrait or what he is seeing, Painting a repro or wat so ever it is ok. You would even hang it in your room. But dare if the Painter would do it in 3D Poser, you would scream " Theft, Ripp, read guidelines " even that the creators have built it up from scratch and just took a template of what has been seen using tools that are given, just like the Painter uses Paint and pencil.

So asking you, is this mentality going to bring you any further? are you expecting that one who might have some advices that could bring you any further will really be willing to share ?  On my side I realized again that it is not the time jet, may be sometime in the future but after all these years I sure doubt it ever will happen.