I just discovered Mage Space, and it's arguably the best free/unlimited generator. You can choose different aspect ratios, and generate a high res version after creation. They're also the only engine to accept mature prompts and generate them accordingly
Lol, I tried a couple just to check that out and I wouldn't recommend it. The engine has no concept of what a person is, so you get a Lovecraftian chimera of porn limbs and torsos. The problem is that it's too incoherent to even be properly disturbing, and just gets boring after you've done a couple.Originally Posted by Azaran
They're also the only engine to accept mature prompts and generate them accordingly
Even when you are not trying for sexualized imagery it adds it. Using the words witch, moon, lithe, woman, beautiful, still produces such. I assume it is because the program has been trained that way by acceptance of downloads done as success. It adds an extra limb or crosses the eyes in hilarious ways but it's still better than it was. It is refining what it observes we want through a self correcting algorithm.
It ignores some things you type such as moon in favor of what it thinks is key despite them being first in order. It has a more lascivious bent due to downloads. It's tempting to blame it's jaundiced eye on the program but the fault is our own. The mention of full body to correct a half or closer image is misinterpreted.
The other program is more artistic but this one produces more realism and sometimes almost gets what you want. Still not as good as painting something yourself but perhaps with time it will listen to implicit instruction and generate a thing exactly. Is it art? In a way. And in another a rehashing of previously accepted imagery. It still produces pleasing works on occasion and leaves you feeling that if only you could find the right words you would have something approaching art.
I did a handful of very specific ones (particular acts described in detail), and about half were right on the money
My bold prediction: within the next few decades, most of our entertainment will be AI generated. The tech to make AI generated music is out there too, and AI video tech is already starting up.
Music, movies, tv shows, will be mainly AI generated within 20 years. You'll be able to feed a full movie /show script into an AI program, and it will spit out a complete film or show within a few minutes. All it will need is human revision to iron out the kinks, some editing, and voilà
That's probably not going to happen like that.
It's possible to pump billions of images into a deep learning engine, because we do in fact have billions of sample images. But ... the search space for "movie" is vastly larger than the search space for "image". The problem is that there just aren't enough movies ever made that would allow the same trick to work. If you need 1 billion images to make an image generator work, you probably need like 100 quadrillion possible movies to make the same trick work for that. The limit of data-driven generators is that you need the data to start with.
And we already have something that solves most of the problems a movie-generator AI would need to solve: video game engines. These don't have issues like forgetting how many fingers a human has, or what was happening in the story 2 minutes ago. Anyway, every second of footage you see in a video game is already generated by "AI". Games already let you tell your own story.
What we've seen for the last 3 decades is ever-more CGI/AI generated content in games and movies, but the costs and amount of labor has always gone up, not down. So the most likely thing going forward is that, yes, more content is generated, but at the same time, humans will be making more than before, not less.
Last edited by Cipheron; 30th Nov 2022 at 22:39.
A Roguelike is kind of an AI generated game each time.
My thing now is generating new paintings by Old Masters. I'm actually considering printing some of these on canvas and hanging them up on my walls
On behalf creative people everywhere, I hope you're right
Last edited by Azaran; 1st Dec 2022 at 11:24.
Let's see it do this-
Painting by Jere Allen, father of an old friend.
It may recreate but it can't original. Its database is still stuck on existing work.
This was a tricky puzzle chamber. To get the tetris piece, you have to get the jammer fr... Oh wait, wrong game.
I had a good run with this Impressionist Hot Air Balloon series.
I've seen AI do pretty remarkable remixes - such as Jodorowsky's Frasier - which I'd consider relatively original. Not wholly, obviously, because they're remixes, but then, I believe that most new art is a remix of various influences to a large extent.
It can create pretty original works emulating the styles of existing artists. If I came across these, I would assume they were original Leonora Carrington paintings
I imagine attempted art fraud will grow rampant, with people showing off 'long lost' works by famous artists
Last edited by Azaran; 2nd Dec 2022 at 11:26.
And the search space for "game" is vastly larger than that of "movie".
I think you might be right. Although I would argue that a generator doesn't always need to be data-driven in this specific way. There are many ways to use neural networks, and many ways to connect them with other mechanisms.
A movie differs from an image in that there are deeper concepts to be understood in a movie, such as plot and the whole temporal dimension. I'm not an expert on AI, but if neural networks are a mechanism for measuring concepts, and that they can be chained to measure higher-level concepts from lower-level ones, then for movies we'd need longer chains of these and thus more compute power.
Another way I'd approach this is that it wouldn't need to learn entire movies, and it wouldn't need to be one chain of neural networks but rather separate engines. There could be an engine that understands space and lighting, and one that understands short-term and long-term causality, one that understands human facial and body expressions, one that understands psychology, one that understands how plots work in textual form, etc.
So in this case we'd give it movies to learn visuals, not necessarily plot. Storytelling it would learn from movie scripts combined with storyboards so it gets the connection how the script translates to something visual.
Yes that last point is how I'd think about it too. As a technical point, the data works at different levels of overlapping scale. For a movie, you don't need a quadrillion images. You need additional levels of scale e.g., starting at the top level with the broad plot arc, including characters, motivations, and setting, etc., then all that decomposed into a series of scenes, then that decomposed into a series of actions with their own mini-arcs for the scene, and only then creating the setting and action, probably not directly via images but rendering models in a 3D engine.
And I do think AI will be making very original-looking art just because I think the conceptual space it's working with is much bigger than our conceptual space of what's "recognizable". The issue isn't the AI's capacity. I think it's the ability of prompt-writers to understand the model and know how to pick the right prompts and prompt-logic that works with the model to create interesting and original-looking art. When you significantly change the model, you can change the prompt-logic.
There's a learning curve happening right now where people that are really into this are learning this new language and getting better and better at it. One of the interesting threads in the SD Discord I thought was the Chinese Telephone game, where a person tries to recreate the image above theirs. It's like watching this community learning how prompts work at a really fine level of detail in real time. What they're really doing I think is feeling their way around inside the guts of the model and how things are "organized" in there, and pulling structure out of it. It's not always intuitive in natural language at all.
So it's not only the model, but it's the users that are advancing in leaps and bounds how to use this tech, at least it seemed to me.
Even printed on canvas there would still be the problem of brush strokes. As a fake in someone's house? Sure. You could use some clearcoat to simulate brush strokes. But under the eye of anyone with training? Nah. I could see a well trained artist duplicating the duplicate and passing it off but that isn't anything different than what they do now. And if the works of an artist aren't already in the database then what it comes out with is nothing like the artist. I put in Jere Allen and it came back with crap. Granted he isn't in the top 100 in terms of famous but here and in Germany he has done well.
A person paints what the AI produced. I don't think the issue will be so much about fraud, but I do think people would be interested in buying painted AI versions of their favorite artists, even knowing it's AI because it looks good, and that will do something to the market. Maybe not the billionaires and banks interested in flipping them, but just normal people interested in art.
While we're predicting the apocalypse of AI art, check out what's been happening with AI chat lately.
This is from ChatGPT.
Insane. I've asked it to provide new hymns to historical deities and it sure delivers. Even adds coherent rhymes. This is a 'new' hymn to the moon goddess Selene:
"Hail to thee, Selene, shining goddess of the night
Your beauty illuminates the dark with your gentle light
You ride across the sky on your silver chariot
Guiding the stars and the moon, a celestial pilot
Your radiance is a beacon, guiding us through the night
Your face is a reflection of the sun's warming light
You are the bringer of dreams, the mistress of sleep
Your touch is gentle, your presence a soothing keep
We offer you our praise, O Selene, divine queen
May your light continue to shine, ever serene
Bless us with your beauty, and grant us your grace
As we journey through the night, guided by your face.
May this hymn to Selene bring you peace and calm
And may your light always shine, illuminating the night's dark realm."
PS: I've googled extracts from these and come up empty, which means the AI is actually thinking and creating them
A hymn to Cthulhu:
Great Old One, Ancient One
Cthulhu, hear our plea
In the sunken city R'lyeh
You sleep beneath the sea
Your power is beyond measure
Your might we cannot see
But we know you are waiting
For the day you'll break free
From your slumber you will rise
And bring forth chaos and fear
Your reign will be unending
As you reign supreme, oh Dear
Great Cthulhu, Great Old One
We worship at your feet
In your name we make our offerings
And pray for your return, oh sweet
May your wrath be upon us
As we await your arrival
Great Cthulhu, we praise you
And pledge our loyalty, oh vile.
Last edited by Azaran; 5th Dec 2022 at 20:26.
Jerry Seinfeld dabbles in witchcraft. Courtesy of Chat GPT
Scene: Jerry's apartment. Jerry is sitting at the kitchen table with a book of spells in front of him.
Jerry: (to the audience) So I decided to try my hand at witchcraft. I mean, why not? It's not like I have anything better to do.
Kramer bursts into the apartment.
Kramer: Jerry, Jerry! I just saw the weirdest thing!
Jerry: What?
Kramer: I was walking down the street and I saw this woman with a broomstick! She was flying!
Jerry: (chuckles) Yeah, I know. That was me.
Kramer: (stunned) You?!
Jerry: Yeah, I've been dabbling in witchcraft.
Kramer: (excitedly) Oh man, this is great! I've always wanted to fly on a broomstick! Can you teach me?
Jerry: (skeptical) I don't know, Kramer. This stuff is serious business. It's not a joke.
Kramer: (pleading) Come on, Jerry. Just one little spell. I promise I'll be careful.
Jerry: (reluctantly) Okay, fine. But if anything goes wrong, it's on you.
Kramer: (eagerly) Great! What do I do?
Jerry: (reading from the book of spells) First, you need to gather some supplies. You'll need a broomstick, of course, and some herbs and crystals.
Kramer: (excitedly) I'll be right back!
Kramer rushes out of the apartment. A few minutes later, he returns with a broom and a bag of supplies.
Kramer: (holds up the broom) I got the broom! And I found these herbs and crystals at the health food store.
Jerry: (looks at the herbs and crystals skeptically) Uh, Kramer. Those aren't the right herbs and crystals.
Kramer: (confused) They're not?
Jerry: (explaining) No, those are just for cooking and decoration. You need specific herbs and crystals for spells.
Kramer: (disappointed) Oh. Well, I guess this means I won't be flying on a broomstick after all.
Jerry: (smirks) Looks like it.
Kramer: (sighs) Oh well. Back to the old drawing board.
Jerry: (chuckles) Yeah, I think I'll stick to stand-up comedy. At least I know it's not going to backfire on me.
Kramer: (nods) Yeah, you're probably right. (pauses) Hey, have you ever thought about doing a spell to make your jokes funnier?
Jerry: (laughs) Now that's just crossing a line, Kramer.
Kramer: (shrugs) Just a thought. (exits the apartment)
Jerry: (to the audience) I guess some things are better left to the professionals. (closes the book of spells)
End scene.
I chose..
circles keys thief night sky
https://www.dropbox.com/s/mngdkjqhznihq2d/circlekeythief.JPG?dl=0
Colour me impressed, I actually wouldn't mind having that framed.
Some of these in this thread are quite impressive.
Oh dear. I think you’ve misinterpreted the A part of AI.
No offence to anyone, but nothing in this thread so far has really been good art. I mean that Seinfeld bit wasn’t remotely funny. Yeah, so a computer can parse loads of data and come up with an approximation of ‘something artistic’. Unfortunately there is no intent, no real intelligence. The mind part is missing - the most important part of art.
Edit - lest I come across as a total naysayer: it’s kinda interesting, I must admit. And there’s certainly potential for this to spit out interesting triggers, or somewhat randomly generated stuff which a human intelligence could build upon. However, I gotta re-iterate so far most of the images herein are (to my eye) pretty ‘effin cheesy.
Last edited by zacharias; 6th Dec 2022 at 04:13.
That Seinfeld bit is pretty coherent.
Anyway, no one has really posted good art here because they're either still using a weak system or aren't using a good one well. If you want to see the state of the art, check out the Midjourney and Stable Diffusion Discord pages.
https://discord.com/invite/stablediffusion
https://discord.com/invite/midjourney
Coherent maybe, but not funny. I’d prefer something incoherent which had a nugget of real comedy.
Anyway, those images are much better. Almost alarmingly sophisticated. I guess I am just a naysayer though because to me it’s just not art - however sophisticated or pretty the images get - if the only human input was a few text strings. I suppose that’s pretty irrelevant to this thread. On a taste level the images still are a bit questionable. I suppose that’s somewhat irrelevant too.
I think it just annoys me because the execution is being left totally to an AI. In reality art can be exceptional because of either concept or execution, or the merging of both. So to me AI art is outsourcing one half of your creativity. I suppose I am opposed to it in principlePointless really because the general public won’t care about any of these arguments, I must admit.
Keep in mind that 99% of human made art is formulaic or downright copycat, comedy included.
Posting Permissions
Reply With Quote
