Tantabus
Filters (Default) SettingsRegisterLogin
2 2 2 0 33
Subscribe
Galleries
Featured in
Sign in to create a gallery
Related
View VS Download DS
Uploaded by Importer
 2048x3072 JPG 3.14 MB
Description

Image imported from Derpibooru.
Original: https://derpibooru.org/images/2991681
Metadata at the time of import:
Field Data
Favorites 162
Upvotes 223
Downvotes 13
Score 210
Comments 33
Uploader DexterMorgan
Original description below this line
Soft: Automatic1111’s Stable Diffusion webui, AstraliteHeart’s pony-diffusion-v2 model and 4x_foolhardy_remacri as upscaler.
Prompt: Fluttershy wearing a hoodie, anthro, cute, detailed, high res, 4k, trending on art station, art by greg rutkowski and alphonse mucha
Negative prompt: ugly, morbid, extra fingers, poorly drawn hands, mutation, blurry, extra limbs, gross proportions, missing arms, mutated hands, long neck, duplicate, mutilated, mutilated hands, poorly drawn face, deformed, bad anatomy, cloned face, malformed limbs, missing legs, too many fingers, extra toes, missing toes,(sfm)
Steps: 20, Sampler: Euler a, CFG scale: 12
And a bit of removing artifacts with photoshop ;)

Tags
Edit
safe
13611 ai content
38839 ai generated
37888 automatically imported
33020 derpibooru import
33202 edit
564 editor:dextermorgan
1 generator:purplesmart.ai
4581 generator:stable diffusion
20522 fluttershy
3377 anthro
13249 g4
32063 backlighting
152 clothes
14216 cloud
1366 female
32725 high res
3377 hoodie
491 jacket
522 sky
1338 smiling
12560 solo
28447
Report Share List favoriters

Comments

Edit Preview
Syntax quick reference: **bold** *italic* ||hide text|| `code` __underline__ ~~strike~~ ^sup^ ~sub~

Detailed syntax guide

33 comments posted
Background Pony #6541
@Background Pony #6541
All the time I see great art not being uploaded simply because nobody has the time to actually upload it. I try to do my small part with uploading, tagging and everything, but it’s barely a dent. Artists uploading and tagging their own art is certainly very helpful. They have the original resolution images, so I don’t have to coax DeviantArt and Twitter into giving me the least compressed version. They know the OC name, so I don’t have to spend 30 minutes reverse image searching and digging up dead askblogs from the Internet Archive. Every little bit helps.
TL;DR If you have time, please upload and let the upvotes, faves and comments be praise. :)
Imported from Derpibooru
Posted Report
Background Pony #6541
@Background Pony #6541
Maybe someday. They were on ponychan, and my feeling was to not post them to ponibooru myself, but instead if people liked them enough, other people could (and I’d interpret it as praise.)
Imported from Derpibooru
Posted Report
Background Pony #6541
@Background Pony #6541
You have misinterpreted my viewpoint. I’m arguing that the issues at hand are human ones.
By pointing out how easily the same arguments work for other categories of art, ones that are generally accepted (with varying perceived value, based on peoples’ tastes), I’m demonstrating that the issues are not inherent to AI, and should not be framed as such. They relate to how people use it.
I’d even entertain the argument that “people cannot be trusted to use AI image generation responsibly, and we do not deserve to have it available to us”. But all the criticisms people think are arguments against AI image generation itself just don’t hold up.
There is a word for the behaviour you are falling victim to. It is called dehumanisation. You are dehumanising other artists and myself by shrugging off and invalidating our concerns with AI art.
That’s only half right. I’m saying the concerns a lot of people are saying are AI issues are actually problems about how people use it, not problems with AI image generators existing. I’m saying you are criticizing the wrong things, undermining yourselves.
With the dangers it presents to our carriers; Skill sets some of us spent a life time honing. The irony is in doing so, you are the only one here who’s acting inhuman.
It’s easy for me to dismiss jeopardizing work, because I’m comfortable. But I think the “muh jobs” viewpoint is harmful in the long run.
Give it a year. There will be companies replacing game concept artists with AI. There will be AI art advanced enough to completely replicate an artist’s style.
Yes, this appears inevitable.
Look up at a city glowing with 100 pictures and become incapable of knowing whether any of them were thought about or created with intent or feelings.
Whether you think there somehow is something desirable about a current-day stock photo on a billboard, or that people will use an AI generated image without even glancing at it to see if it works for whatever purpose they have (which absolutely counts as “intent”, albeit the bare minimum), this isn’t an AI issue.
when you conflate apples to oranges it feels to me as though you’ve never valued the time committed to art. That you’ve never considered or appreciated what makes AI able to draw good art in certain styles in the first place. That you yourself have never drawn a single thing for longer than a couple minutes. You come across to me like a selfish bad actor, like a choosing bagger.
I don’t make money through art, but I’m definitely an artist, with some hundred or more pony drawings (most not on Derpibooru) and probably a thousand in other topics/media. I prefer to stay anonymous, so it doesn’t matter and I’m sure you won’t believe me.
To you the difference between an AI computer algorithm and A living breathing person who wishes to communicate something special through a medium are indistinguishable.
To an extent, yes. People can creatively use any tool and medium - a chainsaw, a toothbrush, a broken piano. People can creatively use AI image generation and to say otherwise is the true dehumanizing perspective.
As I wrote, I take issue with people putting in minimal effort and wished for people to hold what they make to higher standards.
You re-word the points I reflected on and though about by turning around my points as though My living breathing ass that relies on commission money to thrive is in any way 1 to 1 comparable with an AI Promt-crafter who feeds an artists style through an algorithm then publishes a composition after 5-20 iterations.
Absolutely comparable. That doesn’t mean equal, but it does mean both have worth.
With AI art its the opposite. People gawk at something that took 20 minutes to create and didn’t require thought or art skills to do so and the prompt crafters who create it are not incentivised to get better at art. In fact nobody is, as many artists get demotivated when they see a line of code that can do what they do, in their style, only quicker and better and get more engagement.
You really don’t think there’s any competitive drive? You don’t think people desire improvement? People already are discussing prompts (and the many other factors and techniques that affect their ability to realize their vision with AI image generation), comparing results, looking for better results. People already are experimenting with new and creative ways to use these tools to make surreal animations, tell stories, and get inspired. How inhuman are you that you don’t believe in these characteristics of people?
Imported from Derpibooru
Posted Report
Background Pony #6541
@Background Pony #6541
You seem to think that AI will replace artists. It won’t, unless it gains consciousness and becomes a person.
All the same arguments were made about photography. A master painter spent their whole life learning how to paint, then spent days to create a beautiful portrait, expecting to get paid for all this time, effort and expertise… And then some schmuck comes in and takes a photo for a dollar. O tempora, o mores! Painters will surely starve and die, only photographers will be left!
But the thing is that painters did not disappear. Because photography is not a replacement for painting. It can replace only some use cases of painting and not even fully. It even helps painters be better painters - you can use photos as references for painting. You can look at a photo of a famous painting and learn something from it. You can take a photo of your painting, show it to people across the globe and ask for critique. Photography is just a tool.
Similarly, AI is not a replacement for painting. It’s just a tool. It can and will be used by digital and perhaps even traditional painters. You might for example use it to quickly mock a certain composition for a commission and ask the customer if they like it. You might use it to generate a smooth animation from keyframes. You might use it for coloring sketches. You might use it as an inspiration to break out of an art block.
BTW did you notice how AI generated images usually look weird? Either anatomy is fucked, the composition is shitty, the details don’t match or even the entire image is an abomination. Even on images with tons of manual inpainting corrections. And people somehow still upload them. They don’t even notice the issues. They lack the eye of an artist, so they cannot use AI to its full potential. Painting, photography, dreaming up images with AI… The tool does not matter, it takes an artist to create quality art.
Imported from Derpibooru
Posted Report
Background Pony #6541
@Background Pony #6541
What you’re describing on the other hand is an obsession with human creation. It is irrational and really not something you should hold on to when better options present themselves. Technology’s purpose is to replace humans by amplifying tasks previously done by many humans or outright replacing them all together. Things like robots have been replacing assembly line workers as a single robot and a small amount of people to maintain it can do the work of 100, or a combine machine can do the work of 1000 farmers, all of this is just vital to the speed and efficiency of the modern world. There is no reason to hang on to humans doing these jobs as it simply would be wasting resources and really just make the world we have right now impossible (unless you want the population to have to go back to 90% of people being farmers to even provide basic things like the ability to feed yourself).
This of course will happen to all fields though and has been happening all the time. Things like digital art programs or 3D modeling/rendering software allow for much more sophisticated creations of artwork by far fewer people or just help to accelerate what an individual artist is able to do. Now of course the potential to automate even more of the process is coming, but it’s still not a replacement for the high level human creativity and flexibility that makes art valuable as you’ve noted, which is why humans still are involved in the process (for now), but that is just how automation goes. Some day AI in theory will replace us all since we’re automating humanity away incrementally but that’s not really worth worrying about since it’s a far off thing which we may never truly reach. For now it’s better to just adapt to the times and use the new tools provided to you rather than try to hold on to an inefficient past for the sake of humans.
Tldr, You’re not going to stop the advancement and utilization of technology. You’d be better off telling us how we should move back to rooms people calculating things by hand as computers stole jobs away from all those people. I think something you’re missing really as is evident by computers and technology is in general is that new things like this produce new fields and new jobs for people. Programming was hardly a thing back in the day and now it’s a huge field due to the need to make machines work, same with engineers and everyone else involved in building a robot. AI itself has provided tons of ML jobs to people, and this transition to more AI-based art will likely improve art’s accessibility and no doubt have other impacts we can’t really predict on the overall way things are done in the world. Will it annihilate some types of jobs in the process (or at least make them into a more niche hobby thing)? Yes. Is that a bad thing? No.
Imported from Derpibooru - Posted by LemonDrop
Posted Report
Background Pony #6541
@Background Pony #6541
There is a word for the behaviour you are falling victim to. It is called dehumanisation. You are dehumanising other artists and myself by shrugging off and invalidating our concerns with AI art. With the dangers it presents to our carriers; Skill sets some of us spent a life time honing. The irony is in doing so, you are the only one here who’s acting inhuman.
Give it a year. There will be companies replacing game concept artists with AI. There will be AI art advanced enough to completely replicate an artist’s style. Look up at a city glowing with 100 pictures and become incapable of knowing whether any of them were thought about or created with intent or feelings.
when you conflate apples to oranges it feels to me as though you’ve never valued the time committed to art. That you’ve never considered or appreciated what makes AI able to draw good art in certain styles in the first place. That you yourself have never drawn a single thing for longer than a couple minutes. You come across to me like a selfish bad actor, like a choosing bagger.
To you the difference between an AI computer algorithm and A living breathing person who wishes to communicate something special through a medium are indistinguishable. You re-word the points I reflected on and though about by turning around my points as though My living breathing ass that relies on commission money to thrive is in any way 1 to 1 comparable with an AI Promt-crafter who feeds an artists style through an algorithm then publishes a composition after 5-20 iterations.
You openly admit to “”“indulging””” in the whataboutism fallacy as though it does something for you and doesn’t just make you look thoughtless.
You seem to be trying to point out a double standard by conflating that I think highly of SFM, bases or YCHs. I don’t. In fact I’ve never conducted a YCH purely on principle even though sometimes its more lucrative.
In my opinion YCHs are less valuable than original pieces. You pay a smaller fee to stick your OC into a cookie-cut scene and situation along with 20 other people all doing the same thing. There’s less value in that. That’s why YCHs are typically cheaper.
You raise SFM art as a point and again, while the animation, if done well, has artistic merit, I agree in that a don’t see much value at all in slapping some premade assets into a team fortress map and making them fuck. Still SFM art is truer to the idea of a real artist communicating a real idea to a real consumer than AI art is.
Lastly, yes base art is also a short cut, but at least its undertaken by young people trying to get better at art and typically isn’t taken too seriously. With AI art its the opposite. People gawk at something that took 20 minutes to create and didn’t require thought or art skills to do so and the prompt crafters who create it are not incentivised to get better at art. In fact nobody is, as many artists get demotivated when they see a line of code that can do what they do, in their style, only quicker and better and get more engagement.
Imported from Derpibooru - Posted by Wispy Tuft
Posted Report
Background Pony #6541
@Background Pony #6541 replace “AI art” with something else and it’s still true
How would the AI be nudged towards “good art” if it did not compile and examine the pompted artists work? Arn’t trained weights trained via people affirming good art and discouraging bad art to the AI? Whouldn’t that involve the AI having a jumping of point to start at in which it referenced and amalgamated peoples art with an image to image process? How would an aspiring artist be nudged towards “good art” if they did not compile and examine established artists’ work? Arn’t a person’s sense of aesthetics trained via people affirming good art and discouraging bad art on social media, etc? Whouldn’t that involve the new artist having a jumping of point to start at in which it referenced and amalgamated peoples art with an observe-and-adapt process?
This shit is beautiful and haunting, but so fucking meaningless. A line of code spit it out with 19 other pics that look functionally identical to it. It was not thought about or felt about. It was simply made. What value is there in 20 prompts and a machine doing all the interpretation and expression? YCHs can be beautiful but are so fucking meaningless. A dude with a tablet spit it out with 19 other pics that look functionally identical to it. It was not thought about or felt about. It was simply made. What value is there in 20 OCs and someone who needs money doing all the interpretation and expression?
Then the art-sites swell with it and engagement floods to the AI art and people who spend hours drawing picture get ignored. This is frustrating. It fucks over artists Then the art-sites swell with it and engagement floods to the Source Filmmaker art and people who spend hours drawing picture get ignored. This is frustrating. It fucks over artists
I would compare AI artists to kids with coloring books. The picture is already provided, all the artist does is touch up the picture after and then call it their creation. I would compare base-edit artists to kids with coloring books. The picture is already provided, all the artist does is touch up the picture after and then call it their creation.
Yes, I’m freely indulging in the fallacies of “whataboutism” and “more than one thing can be true, even if we’re only talking about one thing right now”. But come on. These aren’t arguments against AI art.
The only problems I have with AI art are behavioral problems of the people using it, and aren’t even unique to AI art either, just exacerbated with it.
  • It can be hard to tell whether someone had a vision and worked hard to refine and achieve it, or just got lucky on a whim - hard to tell how highly to regard the effort and creator’s skill (unless they go out of their way to prove it, eg. by recording a timelapse of the whole process, which shouldn’t be required). …but the same goes for other art, too - did someone use a premade background? Did they model their 3D objects from scratch? How lightly or heavily did they lean on custom brushes that someone else made?
  • It can create a large quantity of, on the surface, nice-looking pictures which people might upload in excessive amounts. …which is something people working in 3D sometimes end up doing, too. People, please have some standards.
  • It can be used to (morally, if not per law) plagiarize people, either by maliciously fine-tuning a model to very closely recreate someone else’s art (a trait called “overfitting” and considered unwanted by researchers), or by using the “image-to-image” functions to steal a pose and make derivatives. …as if tracing wasn’t always a problem in art.
Imported from Derpibooru
Posted Report
Background Pony #6541
@Background Pony #6541
If you’re concerned about low effort content flooding art sites then you should probably focus more of your energy on screenshots which have 6x as much posts as the AI category over the the last month.
It is amusing to see people regress through 100 years of art theory just because they’re bothered though. Are we going back to the era of counting brush strokes to determine if something is ‘real art’?
Imported from Derpibooru
Posted Report
Background Pony #6541
@Background Pony #6541
I mean yes I agree, its low effort art, not that time doesn’t go into AI but I spend days on some pictures you know?
For example this picture has some things that could be touched up
-Her hair doesn’t flow into and connect with the scalp
-the hood of her hoodie flows into her undershirt instead of the hoodie
-Her eyes could be scrapped and redrawn
-Her left ear is fucked up and her right ear is missing
I would compare AI artists to kids with coloring books. The picture is already provided, all the artist does is touch up the picture after and then call it their creation.
Imported from Derpibooru - Posted by Wispy Tuft
Posted Report
Background Pony #6541
@Background Pony #6541
Go back to #Lemonaids
I understand there are multiple steps and iterations as well as touch ups done with AI art. Thats swell. Then the art-sites swell with it and engagement floods to the AI art and people who spend hours drawing picture get ignored. This is frustrating. It fucks over artists
Imported from Derpibooru - Posted by Wispy Tuft
Posted Report
Background Pony #6541
I sure love watching people insulting people for doing what they enjoy.
Imported from Derpibooru - Posted by Butters Stotch
Posted Report
Background Pony #6541
@Background Pony #6541
Art can most definitely be typing commands as programming can result in many kinds of computer-generated art just the same by inputting text and having a computer output something pretty. It’s merely the injection of human creativity and vision into the process which makes it art as opposed to just random output from a totally automated process. I am not interested in your restrictive self-imposed definition of art really, you are not the one who gets to gatekeep it.
Imported from Derpibooru - Posted by LemonDrop
Posted Report
Background Pony #6541
@Background Pony #6541
Digital art= the artist directly inputting their movements and techniques to create a picture piece by piece.
AI “art”= type in commands until it comes out the way you want it.
Don’t bother with the false equlivency, bud.
Imported from Derpibooru - Posted by AwkwardLuna
Posted Report
Background Pony #6541
@Background Pony #6541
No, a human created and tuned prompts to match their vision using AI as a tool to produce the output, it is art. If you doubt this remember that all digital art was rendered by emotionless machines too, unless you think art programs just sit there doing nothing.
Again, if you think it’s “easy” to produce good quality AI art like this, just go try. You’ll quickly find it is not and requires a lot of effort usually to do.
Imported from Derpibooru - Posted by LemonDrop
Posted Report
Background Pony #6541
@Background Pony #6541
If you think this isn’t art try to make it yourself and you’ll find that it is not as easy as you claim to actually make good looking AI art like this. There’s a process and time involved with creative vision from a human and that makes it art just like any other sort of medium someone can express something in.
Imported from Derpibooru - Posted by LemonDrop
Posted Report
Background Pony #6541
@Background Pony #6541
@Background Pony #6541
How would the AI be nudged towards “good art” if it did not compile and examine the pompted artists work?
Because of how CLIP works. CLIP takes text and images and maps them into the same (768 dimensional) space. Points nearby in this space are similar in both semantics and style. There are aspects of this space that impact art quality. You could train Stable Diffusion without ever even seeing a single work by Greg Rutkowski and prompts using his name would still improve image quality given a CLIP model trained on his work.
Arn’t trained weights trained via people affirming good art and discouraging bad art to the AI? Whouldn’t that involve the AI having a jumping of point to start at in which it referenced and amalgamated peoples art with an image to image process?
See I’ve had like 5 people tell me AI doesn’t reference real photos and real peoples art yet when you cut through the crap and the bias training where the AI figures out what looks good and what doesn’t, you get real references.
Trained weights are created by looking at existing images. But these weights do not “contain” these images. The weights contain what the AI learned from looking at the images. Referencing is done by looking at another image while creating something, the AI can’t do this. You can ask it to reference Castle by Greg Rutkowski all you want and it has no idea what that image is.
This is analogous to the difference between being inspired by a lifetime of looking at the world, art, tv, movies, etc… vs. referencing a specific image or set of images.
This shit is beautiful and haunting, but so fucking meaningless. A line of code spit it out with 19 other pics that look functionally identical to it. It was not thought about or felt about. It was simply made. What value is there in 20 prompts and a machine doing all the interpretation and expression? This picture has no artist, how can it be art? Its like calling a beautiful lush valley art.
I’m not going to argue about this part as it’s really just opinion. People like what they like and see meaning in different things. I see no issue in people finding AI art meaningless, I just care that people aren’t misrepresenting how these systems work.
Imported from Derpibooru - Posted by Bigcheese
Posted Report
Background Pony #6541
@Background Pony #6541
AI is just a tool. If all the artist does is make a prompt and then lightly edit the result, it’s low effort art, but still art.
Consider this famous painting: https://en.wikipedia.org/wiki/Black_Square_(painting) - it’s still art, despite being low effort.
When photography was invented, people were similarly lamenting that it devalues art, that it’s soulless and that artists will now starve because any fool can take a photo and easily create a very realistic image.
Imported from Derpibooru
Posted Report
Background Pony #6541
@Background Pony #6541
How would the AI be nudged towards “good art” if it did not compile and examine the pompted artists work?
Arn’t trained weights trained via people affirming good art and discouraging bad art to the AI? Whouldn’t that involve the AI having a jumping of point to start at in which it referenced and amalgamated peoples art with an image to image process?
See I’ve had like 5 people tell me AI doesn’t reference real photos and real peoples art yet when you cut through the crap and the bias training where the AI figures out what looks good and what doesn’t, you get real references.
This shit is beautiful and haunting, but so fucking meaningless. A line of code spit it out with 19 other pics that look functionally identical to it. It was not thought about or felt about. It was simply made. What value is there in 20 prompts and a machine doing all the interpretation and expression? This picture has no artist, how can it be art? Its like calling a beautiful lush valley art.
Imported from Derpibooru - Posted by Wispy Tuft
Posted Report