Byrdie posted at 10:23 AM Wed, 19 October 2022 - #4447004

Since I use Poser, Daz, Photoshop and now Stable Diffusion AI to get the results I want, mixed medium is probably the best place. Mixed is a good fit, imho. 

i wouldn't consider it mixed media either if i used an AI image (or as i did once, a selfmade terragen render, or any input from another app like fractals) in a scene where it doesn't really matter and is small-ish, like a poster, a framed painting in a DS scene/render, because it doesn't really change the weight of the DS 3d scene per se.
but i guess i would consider it mixed media if it was a full blown background (from another app or especially here AI) that really influences the scene tho, since this would weight a lot on the result. after all, a DS 3D render is a DS 3D render, where i get my textures from is irrelevant - otherwise every single texture from a photo would be mixing media, in a way.

so, what do you folks think about "correcting" AI works? i mean like, repairing hands disasters and the likes? because since AI art is mostly made by well, the AI, so the skills are in the prompting, directing the AI, and choosing the renders.. is it "cheating" to correct unsightly bits (and then collect the viewers applause for a "perfect" render)? is it expected that you declare the postwork even for smaller corrections? should it be considered mixed media because gimp etc were used? even if in the end effect, what you have/exhibit is.. an AI generated work, in the 1st place.

gawd, this was a slab of a comment. sorry, being succinct is NOT one my talents, lol...

Man, artists are really being hit by this.

https://lesbiacebian.tumblr.com/post/703326410379591680/i-dont-know-what-else-to-say-except-that-ai-art

There's also AI being trained to write fiction.

Maybe we'll all be replaced by robots.

today yesterday or tomorrow is Proclamation Day here (I am unsure which) the Holiday is today 

Great! This is a mega solution. I run my own business and generating graphics using AI is something I've been waiting for. If you also run your own shop then I recommend reading the article on smart packaging: https://authena.io/smart-packaging/. It's a very interesting solution.

Wow. Hm. I seem to remember being scoffed at about this subject not long ago.

Looks like certain commerce corporations are also embracing intellectual property theft by providing a home for "artists" that endorse such theft. Ironic. On principle, by example, I would say this indicates that it's perfectly fine to steal other's art works. For instance the models that are sold on this site and can be found on various GFXer sites. Is there a difference?

It's perfectly ok then? One should feel at liberty to steal and share the work of others right? I mean the actions here are an endorsement to do so right?

Ok... was just trying to figure out the moral liberties of such a thing. After all, an eye for an eye I guess...

I trained an embedding using my ugly face

and I hope the user blocked from my view cannot see them because while I don't recall the username I don't block people lightly or they blocked me 🤣

Video using my embedding

I don't have a Midjourney Subscription and refuse to pay for ai art as it now stands

it has no legally endorsed commercial use or ownership protected by law as well as questionable ethics however the Laion 5b database does cover educational and developmental use which falls under my usage of it

(maybe if a model was released trained on public domain images and stuff used with permission I would open my wallet)

so that is why I tend to post in this thread

this used the Stable Diffusion Webui in Visions of Chaos that loads the Automatic 1111 repository

Emma Nippon-Stone 

And a few other versions of her 

(Stable Diffusion)

Boy, she gets pretty busty as you scroll down the page.

I used https://www.mage.space before which was entirely free till like a week ago, it turns out they all want money. Don't pay for any AI on any website, they WILL sell your info. 

hborre posted at 11:02 AM Sat, 14 January 2023 - #4453895

Boy, she gets pretty busty as you scroll down the page.

A class action suit has been filed against Midjourney and Stable Diffusion. If your name is on either of these lists, you are eligible to join the class action lawsuit.

https://twitter.com/Kelly_McKernan/status/1615066246591938560

randym77 posted at 12:16 PM Wed, 18 January 2023 - #4454169

A class action suit has been filed against Midjourney and Stable Diffusion. If your name is on either of these lists, you are eligible to join the class action lawsuit.

https://twitter.com/Kelly_McKernan/status/1615066246591938560

I am not a lawyer but surely  there has to be some sort of proof of loss or earnings, damage to reputation etc

finding billions of image files in a 4GB.ckpt is going to be quite the challenge too

I am not saying it was morally right to use LAION 5b to train the model but making a legal case against it is going to take a lot more than saying you shouldn't be looking at my publicly posted jpg online (especially as it wasn't rightclicked and downloaded even just the thumbnails with description in the Google search engine trawled basically )

I have no idea what the standard of proof is. Artists are complaining that it's affecting their income, but whether they can prove it is another story.

Though I guess if they can prove some kind of copyright violation it would be easier.

one and only time i will  comment on the subject..... a very hard nut to crack....from what i have read on the subject....

LAION-5B: An open large-scale dataset for training next generation image-text models...was as far as i'm aware....reviewed by a extensive group of AI ethics researchers..on a academic and scientific level before its release date,

AI art models "do not store com­pressed copies of [copyright-protected] train­ing images” and then “recombine” them; functioning as “21st-cen­tury col­lage tool[s]  AI art models do not store images at all, but rather mathematical representations of patterns collected from these images. The software does not piece together bits of images in the form of a collage, either, but creates pictures from scratch based on these mathematical representations/ approximation...if it was as individuals suggest using images...then it would be a perfect reconstruction, or reproduction....so common sense and logic say its Computer Pattern Recognition....

Copyright law protects finished works of art. It does not protect things like facts, ideas, procedures, or an artist’s style, no matter how distinct. there is no country in the world that copyright law covers an artist style, artistic techniques, artistic genres, or art movements....

There is also that Fair use....doctrine in United States law that permits limited use of copyrighted material without having to first acquire permission from the copyright.....

"The law of copyright is clear that only specific expressions of an idea may be copyrighted, that other parties may copy that idea, but that other parties may not copy that specific expression of the idea or portions thereof. For example, Picasso may be entitled to a copyright on his portrait of three women painted in his Cubist motif. Any artist, however, may paint a picture of any subject in the Cubist motif, including a portrait of three women, and not violate Picasso's copyright so long as the second artist does not substantially copy Picasso's specific expression of his idea."

It's a minefield that will go on for a very long time....and will have no clear cut outcome, unless they can prove it categorically one way or the other...without any gray areas.....

Can you please not use the names Ross Tran - https://www.instagram.com/rossdraws/?hl=en and Greg Rutkowski - https://www.instagram.com/greg_rutkowski/?hl=en in your prompts? These are two great artists I follow on Instagram. Everytime you put their names in your prompts, feels like you are ready to steal their identity or their souls, I don't know. AI Forums here are public and visible to the two artists also. I've informed both of them that their names are being used to generate AI images here.

He Whose Name Must Not Be Prompted -- "You Know Who" -- has already been jettisoned from anything I do with AI because I tend to get better results without it. Also, I do not know who Ross Tran is, therefore I'm not very likely to use them as reference. 

He Whose Name Must Not Be Prompted -- "You Know Who" -- has already been jettisoned from anything I do with AI because I tend to get better results without it. Also, I do not know who Ross Tran is, therefore I'm not very likely to use them as reference. 

Pretty sure I dropped it from the Muppet Prompt last time I ran that one. I'm using Stable Diffusion, so it may look/work different in MidJourney, which I don't got a sub for and therefore never tried it in. But I mainly do cats, dogs, Snapes and Sherlocks, so even if I were to Muppet-prompt for any of those, "You Know Who" would not get mentioned. Wrong style and like I said, much better without. I try to go as nameless as possible with styles, unless I deliberately want something in a classical -- as in way too old to be anything else but public domain -- vein. Like this bit of nonsense:

Come to think of it, I didn't name any artist style in that prompt either. But it's pretty safe to say Leonardo is not responsible for any "Dino Lisa" on a surfboard shenanigans. With or without the rocket launcher. ;-)

if I use artist names, I generally use a lot of long dead painters and illustrators too, my fave being Ernst Haeckel, I will toss in a few more recently deceased sometimes though.
I really hope they create a fully CC0 sourced training model as I think it can be an awesome tool like the synthesizer was to music.

LOL @ wendy... "Yeah... sorry I'm imitating your artistic style but hey, you're like dead and stuff... you can always use the ouija board and let me know if you don't agree."

Style by itself cannot be copyrighted as the law currently stands. Therefore not illegal, regardless if said style originator has been dead many centuries or is currently living. To make it illegal would require drastic changes both legal and political, and even then no guarantee you'd get the desired results in this case. Be careful what you wish for ...

https://reason.com/2023/01/19/dont-let-disney-monopolize-a-i-generated-art/

I think too many people are seeing this as a black and white issue

I am firmly in the grey area in my views

I don't disagree with many of the issues raised by the say no to ai faction

but

I am not going to ignore a pretty awesome technological tool either

nothing I am doing is depriving anyone of a single cent if it comes down to monetary arguments

I am still buying 3D content too, in fact I am buying more this year already than last year

I, too, is somewhat curious in what A.I. generated pictures may, will, or won't influence the artists that earn money with their art.
And I sincerely respect their fears.
For me, tools like Stable Diffusion are nothing more than another toy that won't avoid me to continue with my Vic4's collection.

IMO dropping a series a words in two "black boxes", move sliders, change a value here or there. select a sampling method or another isn't doing art, even though if this implies knowledge, experience and training. It's playing, like any other game.

Simply because one have very few control on the creating process: you just throw parameters and wait for a more or less random response.

But it remains fun.

A.I represent a for bigger danger when people are using them instead of searching and studying by themselves. The human brain needs to train, to learn, fun is only part of its development, even if it's important. The danger even increases when we can find proofs that often their responses look very well formulated but biased.

I fixed that for you!

This is what I used AI for. The original is a Poser portrait render of V4... I don't know if I still have it anywhere so I can show you, wish I did

This is from Jan '22, before the AI spread like a virus that is now

https://www.renderosity.com/gallery/items/3043772/just-some-cuties

Wow 

I fixed that for you!

Cool, thank you

You're welcome, my friend!

Here is my attempt at a workflow with DAZ and Stable Diffusion. 

I used several of my own blender renders to train an embedding. Then I wanted to combine the embedding and a Genesis 8 custom character that I created,who resembles the embedding, essentially using Genesis 8 as an avatar. That way I can pose her, alter the lighting, change her clothing and hair, render in iray, and then use imgtoimg to add the embeddeding. It could also mimic her facial expressions.

Before I've used iray renderings plus imgtoimg without any embedding involved and with a low denoising value, but it was hit-and-miss. Even though I was using the same seed and prompt, changing the lighting, the clothing, or the camera angle would cause imgtoimg to change the face of my character. Additionally, the aspect ratio itself changed the face too. 

Given that I'm using low denoising values and my G8 character already resembles the embedding (albeit not exactly, but real people look differently depending on the lighting), it ought to have worked.

If I had the same seed each time, it should to have producted the same face. Okay, no! SD doesn't know that this is the same person (or that that it is a person at all). When I tried to increase the denoising value, it initially merely combined the two faces, but it eneded up with different combination for each picture,but and it also made some caricatures, blurry images and nightmare fuel. Also, if I adjust the lighting or aspect ratio, I would get a different person. 

I made several attempts to make it work. It was really hit-and-miss. I was able to make it work in the end and keep the lighting, pose, clothing and hair of the G8 while also producing a person who is reasonably consistent and somewhat resembles my character. The process as a whole was interesting and somewhat irritating. It is still far from ideal, but I like promise of what this could do someday.

Even with lower denoising levels, SD managed to mess up the hands and other little things. 

The embedding is not very consistent in and of itself. The images I've shared are the ones that are more reliable. Even when used in SD with just a prompt it will change the contour of her face, the color of her eyes, and the shape of her eyebrows. Sometimes it would dramatically age her or turn her into a caricature.

I used to overlook all things in the beginning. Now looking back at it, I see plenty of nightmare fuel.

The use of AI will affect every single one of us...Yes it's beautiful and yes it's fun but it will destroy every creative industry...these models were built on scraping the entire Internet... Violating the privacy of the public and Violating Artists IP rights. It will destroy human Lives...The future generations of artists are already suffering. AI needs to be regulated and controlled. In its current state it is unacceptable. Every single picture you put into an AI model becomes part of the Datastack. Every image you generate makes AI smarter. Art is a human expression and skill. An AI host is not human... It can not think..express...feel or Create..The huge issue here is that AI was built and trained on mass laundering of human art. Affecting lives and destroying our creative industries so that CEO and investors can benefit. They are using YOU while capitalising on AI to wipe out all human created platforms. Please be mindful of this and not shortsighted to the impact it will have on our beloved creative communities and the future. Thank you for reading and hope that you will think twice before generating more images. Remember... If you support it you will have a hand in the destruction of all creative industries. 

I agree there is much room for improvement with regard to AI and the use thereof. You'll get no argument from me on that front. And if someone does not wish their work added to the training pool they should indeed be free and able to make that choice and have it respected. But -- and maybe this is just me -- every time I hear somebody proclaim "BLANK needs to be regulated and controlled" (variant: "BLANK must be regulated and controlled because DOOM!!!") I tend to get very unpleasant vibes. The sort that make me want to check which era I am living in on the off-chance I haven't accidentally slipped into the past where that was pretty much the mantra of every authoritarian regime. Or worse, fallen sideways through time and space into a Mirror Universe where it's absolutely barking all the time. 

RedPhantom posted at 8:43 PM Sun, 10 July 2022 - #4441020

I've used Wombo before so I thought I'd compare using the same words: Ocean sunset castle and waterfall

Here is night cafe

and here is Wombo

Not sure why images I've put in my post aren't showing up or why I can't seem to edit them afterwards. This is Night Cafe for me:


And  here's some Lexica Aperture AI:

A Steampunk Marilyn Monroe on a steampunk motorcycle from Lexica Aperture AI.

forum can take a while to load them

if you close the browser you cannot edit your post upon return 

it has to be done the same session 

very nice images 👍🏼

wendyvainity posted at 12:42 AM Fri, 7 April 2023 - #4461086

forum can take a while to load them

if you close the browser you cannot edit your post upon return 

it has to be done the same session 

very nice images 👍🏼

Here are some gorgeous portraits I put together especially for you. Hope you like them.

I'll be back with some more!!!

This is what AI did to two amazing artists here:

https://www.renderosity.com/marketplace/vendors/The_Row_House

https://www.renderosity.com/users/gmlab

gmlab happens to be the Artist of the Year 2022

if you can block your posts, surely you can hide the forum you don't wish to see and spare other's your trolling

their actions are their own

if they don't wish to sell on this marketplace I cannot help if my wishlist was halved

I just will buy other people's digital content for my 3D renders instead

cutting off one's nose to spite one's face is IMO a dumb thing, just because I play with AI a bit doesn't mean I still don't use DAZ studio, Carrara and Unreal Engine DAILY!!!!

and buy lots and lots of 3D content

but .... whatever

enough talk from me, back to doing what I enjoy

a video

3D and AI

I stopped buying 3D merchandise years ago, so what some merchant does or does not do has no effect on me. Childish tantrum and trolling also gets you on my own ignore list.

I don't see any trolling here. If you think you are being trolled by another member here report it to one of the Moderators. Just because you don't agree to what someone else has to say about A.I. doesn't give you the right to call them a troll. Everyone is welcome to post their images here as long as they comply with the website's rules. 

Do not eat lol

posting anti AI diatribe in the AI forum art sharing threads is trolling, there is a few threads in already community forum for discussing that sort of thing, it's like posting Poser evangelism in the DAZ studio forum or Carrara is better than Bryce on the Bryce forum.

Yes users can disagree but this is not the thread to be doing so and worse having the people you reply to on ignore to rub it in, again ignoring is fine but then don't reply to their posts if you are the ignorer.

I'm not the one holding a grudge against someone here. I'm not going to join your "alliance" against twistedkid or anyone else here. 

TwiztidKidd posted at 4:11 PM Sun, 16 April 2023 - #4462377

Here are some gorgeous portraits I put together especially for you. Hope you like them.

I'll be back with some more!!!

1) "Femme Au Béret Et À La Robe Quadrillée" by Pablo Picasso, apparently sold for $67,400,000

2) An untitled drawing by Jean-Michel Basquiat, apparently sold for $15,200,000

I have to say I prefer the two you created. But then I'm a non-'creative' ;o)