Cool Pictures from the Mind of a Machine: AI Generated Pictures

Birdjaguar

Hanafubuki
Super Moderator
Supporter
Joined
Dec 24, 2001
Messages
55,045
Location
Albuquerque, NM
In order to keep our "Cool Pictures" thread from becoming full of AI generated images, this thread is for AI pictures of substance, beauty, and imagination. The Pope does not have such a jacket.



Captions are encouraged. Do not spam this thread.
 
At risk of sounding petulant, where images which are half AI and half non AI should go? For instance a human made drawing used like a starting point or perfected by AI.
 
every AI made picture originates with humans . AI unlikely to be perfecting pictures , at least for a little while . AI result corrected by an artist should be here if known to be so , with a note to that end . With AI , the value of humans going down , not even clowns to be necessary soon , let us keep respecting us for a little more .
 
every AI made picture originates with humans . AI unlikely to be perfecting pictures , at least for a little while . AI result corrected by an artist should be here if known to be so , with a note to that end . With AI , the value of humans going down , not even clowns to be necessary soon , let us keep respecting us for a little more .
Nah. AI is ideal for perfecting pictures. It is able to make a perfectly finished job from a quick drawing or even from a sketch made with a few lines, and being as faithful to the original idea as the artist wishes, if he knows how to do it of course. AI is not going to replace artists but artists who aren't able or don't want to use AI are going to be in trouble.
 
28-03-2023 charlie-strife-screenshot-12.jpg


will kindly avoid the question that created the image was the Pope is an anorak or the issues about livelyhoods and stuff of "artists" who idioticly would think that they have a say in the proceedings of governance after the glorious future that awaits the Mankind in so little time .
 
At risk of sounding petulant, where images which are half AI and half non AI should go? For instance a human made drawing used like a starting point or perfected by AI.
Put them here and perhaps explain the story behind the image. :)
 
So wholesale theft of original art, even worse than what's happening now.

There are artists who sell on deviantArt. I have half a shelf in my china cabinet dedicated to apofiss' cat, sparrow, and owl art, not to mention magnets and a couple of throw pillows.

Then a couple of years ago I found that some company had ripped his images off and turned them into a set of cross-stitch kits on Amazon.


This makes me glad that I finally quit Amazon Mechanical Turk.
 

An AI-generated ‘Balenciaga pope’ fooled us all. How much does it matter?​

The viral image of Pope Francis in a puffy white coat points to our AI future — for better or worse​


Look: The pope’s clothes are almost always interesting. They’re either surreal because they’re arcane and generically holy (an old man traverses the world wearing a long robe and matching hat, like Gandalf) or because they’re startlingly contemporary (the same old man also wears a Swatch watch). The very fact that his daily clothes and accoutrements have to be in keeping with sacred tradition can fascinate, too. His leather loafers should be red like martyrs’ blood; the car he rides is often specially modified for him to stand up to greet the faithful who gather to see him.

So when a photo surfaced this weekend, just before the fifth Sunday of Lent, of Pope Francis in a long, white, trendy-looking puffer coat with his traditional pectoral cross and white zucchetto cap, it’s not hard to imagine what happened next: People went wild. “OKAAYYY,” wrote one Twitter user who shared the image. “Ayo. Blessed be,” wrote another. This particular puffer — gargantuan and gleaming, with a cinched waist and imposing oversize hood — landed in that slim Venn diagram sweet spot between “what the pope might actually, practically wear to keep warm on a cold day” and “what the wealthiest 26-year-olds are currently wearing around SoHo.”

The image was completely fake. According to the fact-checking website Snopes, the image was created using the generative AI program Midjourney and later appeared on the subreddit r/Midjourney.

The coat, for anyone looking to Steal the Pontiff’s Look, resembles Balenciaga’s $3,550 Long CB Down Jacket for women as well as Rick Owens’s some $3,000 Duvet Jumbo Peter Coat. Both are black, but one has to imagine that the designers, like the auto manufacturers who make each new popemobile, might allow a few custom modifications if it were Il Papa asking. The fake coat fooled a lot of people — and it fooled a lot of people in the same week that saw fake, AI-generated images of cops accosting former president Donald Trump. Yes, suddenly it seems all too obvious how artificial intelligence could easily be used to create propaganda, how it could easily be weaponized as a tool of destabilization.

But, that said: The Pope Coat Incident makes clear that AI can and will also be used for the equivalent of making hyper-realistic cartoons. For dreaming up fantasy fashion statements, combining any given celebrity with any given clothing ensemble like an infinite set of paper dolls. For creating the photographic equivalent of fanfic. It may have been one of the first true mass AI misinformation events, in other words, but the puffer-pope saga was also … pretty low-stakes.

The Popecoat, then, arrived at a moment of clear and justifiable alarm over AI-generated imagery, and when its realism had advanced perceptibly even from their capabilities a matter of weeks ago. “The meme likely went viral because of the uncertainty about whether it was real or fake,” said Arvind Narayanan, a professor of computer science at Princeton University who studies AI. Because many more people have access to this kind of technology, it will be important for social media platforms such as Twitter, Instagram and Reddit to develop better tools to quickly label misinformation, he said. “It goes without saying that we can never again assume an image is authentic because it looks realistic.”

Deepfakes have certainly fooled people before: a fake “drunk” Nancy Pelosi video in 2019, a Mark Zuckerberg “announcement” about Facebook ads, also in 2019. But the dripped-out pope, created by a 31-year-old Chicago construction worker who came up with the idea while on shrooms, is a reminder that not everything created by AI is made with the intent to pass itself off as authentic. (“I just thought it was funny to see the pope in a funny jacket,” the construction worker told BuzzFeed.) There’s a word, after all, for the depiction of things that aren’t necessarily real: art.
 
So wholesale theft of original art, even worse than what's happening now.

There are artists who sell on deviantArt. I have half a shelf in my china cabinet dedicated to apofiss' cat, sparrow, and owl art, not to mention magnets and a couple of throw pillows.

Then a couple of years ago I found that some company had ripped his images off and turned them into a set of cross-stitch kits on Amazon.


This makes me glad that I finally quit Amazon Mechanical Turk.
It's copying peoples' styles but it's not literally reproducing their works. It's a huge issue right now, where does someone's intellectual property end? A style cannot generally be copyrighted, only the contents. If style could be copyrighted, the first artist to a genre would own the genre. Really, genres could not exist.

And yes it's scummy to steal someone's art to make a product. But if I were the artist, I would be pleased to know someone set up and proved a revenue channel for me to take over :devil:
 
Legally it is a difficult topic. Any human artist learn to make his art studying other artists works too, including copyrighted ones. So? Anyway it only affects to some models like the ones used by midjourney or Stable Diffusion 1.5 (in more recent stable diffusion models they avoid using copyright art and to train the models)
 
There's definitely an unfairness that if someone has a unique style they don't have quite the head start as AI is super fast. On another hand, if people are using an AI to copy your style you are clearly in demand and should ride than confidence to success.
 
while am an idiot , ı do have weird pals who were not , like back in their day . As such ı have heard Artstation's previous owners refused to take part in some experimental thing where they were to provide the imagery . Can't remember whether that site's sale was during the times when Disney was to buy CFC ... Not denying you to have a computer programme or rejecting imagery and stuff provided as output . But the end product was (once) a set of pictures either drawn or taken by an human being . Uses stuff far beyond my understanding to fuse them , but it is not a thing your computer actually does or whatever . One is supposed to support new things , awright to that ... But do not have to support EVERYTHING .

29-03-2023 10mb hard disc 1960s.jpg 29-03-2023 1968 first  mouse some sort of windows.jpg
pr is not always right and claims regularly fall short ... Before saving pictures to read them , these two prove the development and stuff over time . The first is a 10MB harddisc from 1960s . The other is reportedly the first use of windows type management in which you can choose parts on the screen with your mouse to do stuff . Am yet to hear anyone forced to remove HAL from action . The fear is political , with the feebleness imposed on people suddenly becoming uncontrollable , if agents of chaos can buy the same software you use ...
 
Well, time to post some AI images, uh? and with some story behind. :p

Will post one AI generated image taking as base my own drawing, which i painted manually with Corel Painter some years ago to illustrate a sci-fi story a friend was writting:
Here is my original human made work. It was one of several sketches i propossed to my friend, not the definitive one. I made it in 30 minutes or so:

16 x 20 200.png


And here the Stable Diffusion interpretation i just generated using the img2img method in A1111 with RealisticVision 2.0 model:

00030-700597878.png


Quite an improvement, isnt it? And in 5 minutes aftes some attempts. However it keeps the essence of the original drawing, excepting some details like the gun, which could be also kept with some extra work. Of course to reach this degree of completion manually i would need days, and even then it possibly wouldnt look that good.

This is that i meant for AI being a tool for artists. It is the control and coherence that can be achieved using AI that makes it interesting, not some random pic generated with midjourney from a bunch of words, which i personally dont find of much interest, even if the results are very beautiful.
 
Last edited:
Definitely not an improvement but still very cool. If I had to buy a print I’d buy the one you made.
 
Moderator Action: This thread is for posting AI generated images one one's thoughts on them.
 
Definitely not an improvement but still very cool. If I had to buy a print I’d buy the one you made.
TY. Still mine is at a very crude stage and not suited to be published imo. My point is AI can be very useful for artists. In this case i used a generic model and i prompted the AI to make it in a realistic style remembering 60s scifi novels but you can also train the AI with your own finished work and it will finish your sketches as yourself would. It is very useful and pretty funny too.
 
This is that i meant for AI being a tool for artists. It is the control and coherence that can be achieved using AI that makes it interesting, not some random pic generated with midjourney from a bunch of words, which i personally dont find of much interest, even if the results are very beautiful.
Not being an artist myself, I do like the randomness when I ask it to generate a batch of pictures in some styles and how it can surprise me by mixing existing concepts...



... or how it can render very different styles



 
It is amazing indeed. Specially how it can keep the likeness in very different and very stylized styles (Clint Eastwood), even after having been trained with 10-15 images only. That is something that keeps blowing my mind.
 
Top Bottom