Welcome to my first new post, is this made with AI? Is it hand
drawn? Is it photomashed? Is it photoshoped? Is it made with GIMP or
Krita? And how much? Who knows. That's the mystery.
I
would rather keep my workflow private, but I'm now obligated to answer
'Yes' or' No' to AI tools, and if I don't (by default its No), I get
accused of AI anyway.
It's funny that people want to know if something now is made with AI, but before, no one asked if art was made with photoshop.
In
my own opinion, I think everything digital is one of the same level of a
genre, if the art is created on a computer tool rather than outside of
computer. If anything, all that isn't made without a computer should
have it's own label that people can easily be notified about like
"physical source."
AI is still a new digital tool, but
a digital tool nonetheless. So I understand why people want AI to be
labeled. I know more people are starting to appreciate the use of AI,
but I also agree that AI should be it's own category, not because of
hatred against it, but because it's historically fair. Even with
photography, there is a separate category, merit, awards, etc. But as it
is today, AI is still mostly feared, and has yet to be understood by US
copyrights.
(Ai using a grid composition design)
--FEAR, PREJUDICE, & IGNORANCE --
The negative reception to this new genre of AI Art is really based on two parts: the lack of awareness of what is art, and how AI actually works.
AI has already won awards, so I know that
people appreciate the designs; I just think it leaves a bad taste when
people find out AI was entered in a non AI competition. Wrongfully, its common for non tech people to assume an AI image was done by a bot and not a human using AI to make an image. Not that a bot can't do AI, just as AI text can, but there is a clear difference when an automated computer uses a program versus when a person does. It's the night and day difference between the novice AI artist to the Advanced and Experience. So at early
times like these, I want to choose to be private about all my work until
society begins to see AI as a human tool just like photography.
In these early stages, before we can ever address absolutes, the AI question is really a basic debate on the philosophy of Art itself. As an avid AI user, CathryneDelamort, has stated:
"Traditional" often complain about "art" created using AI models. Works created using generative AI are not “art”. Neither is so-called “art” made with computers to artificially manipulate visuals using 1s and 0s. This airbrushing and vector-ization away of hand-made imperfections is killing the soul of what makes true art. Don’t get me started on photographers who are no more than thieves passing off reality as their own creation. You can not use a man-made machine to create art. Therefore photography is not art.
...
The only true art by real artists is painted with bare hands on cave walls using crushed berries, insects, and animal blood as God intended. Anyone too cowardly to stand up to those not painting in caves should be ashamed. Nothing like this has ever happened before*, and this is where real cave-painting artists must make their stand!
(https://www.deviantart.com/cathrynedelamort/journal/AI-generation-is-not-art-949008519)
(https://i.redd.it/xzdmuflk81261.jpg)Which, brings us to our modern age of where we draw the line in what is considered man-made art vs the actual purpose of creating art. If for example, a purest painter wants to argue that using computer tools is cheating, why is the same standard not applied to physical art that uses the same tools? This is the huge debacle of today's purest argument when someone considers how art is made. Why is it considered cheating for digital artists to use computers and editors, and not considered cheating when physical artists use references, photography, outlining, even digital coloring?
If Art itself is not the question, and a matter of whatever tool the artist decides to wield, then the real center of what is art falls entirely on it's purpose: what function does Art provide to humanity?
This philosophical debate can easily be simplified to it's purpose, and not so much for the infinite reasons people use Art as. The purpose of Art is to emulate, reflect, demonstrate, express, an imperial idea. I personally see this expression as a form of creativity, as I state in many of my posts. But this is just my light of the topic. Others can easily make art without ever being creative, new, or different in the most generic way possible.
So, now that we can understand that Art is akin to an "expression," we can start shooting down so many modern arguments of entitlement about how people attempt to seclude the ways people make art today.
The artist of today commonly uses digital tools in some way to manage their art, even create art through only digital tools. This is nothing new just because AI is now popular. AI Fear is really a larger epidemic through modern society that only views technology through the boom of smartphones, apple, and online shopping. Technology in itself has long been used by artists to make art. This is because it's never been about the computer's purpose, but people that manage to use computers to produce Art. Just like 1's and 0's, if there's a will, there's a way for people to play around with computers to produce and distribute Art.
--NARROW MINDED & ARROGANCE --
If I could summarize the modern age today, it would be a crowd of blind protestors. People that are quick to do away with something new without the knowledge behind what it is or how it works.
As another user yajurka has stated:
"I think that main issues is that since we call our algorithm "Artificial Intelligence" that it has mind of its own and can magically do whatever we ask from it, when in reality it's just lots of statistics and mathematical operations cleverly applied over matrices."
This is all too typical in our internet clickbaits and social media groups. People are quick to hate on AI through their ignorance in how AI works and how AI is applied by people.
It's one thing to understand how AI code is trained, but entirely different to how AI is commanded by people to work. This failure of people to recognize that AI, just like any other computer program, is a tool that humans control, will be our own self harm. The petty surface level complaint that all AI art is stolen just because of how references were used to train the program to function, is a real low ball nitpick to argue. It's like hating a camera on a shopping mall for scanning the faces of people in order to understand how to recognize their human structure and behavior. The point is, the subject is not special and the results are not about the subjects--they're about the development. Or, maybe this is just an attack on open source AI databases and not
privately owned AI, like Adobe?
Therefore, I'm
truly sorry that illustrators feel the most threaten about AI, but the reality is --get over yourself; AI isn't about copying you're physical or digital art, it's about creating a new art genre in the aesthetics of generations. No one is special, no one is a targeted victim.
Like Cathryne points out, "The image is not a copy of something that exists already. It is an amalgamation of things the AI model has been trained on."
If you put a hundred pictures of a house, into a metaphorical blender, how likely is the end result going to look like the referenced original? If AI was a database of human voices, and you placed a hundred sound recordings together, how likely are they to sound like a singular person apart from what a human sounds like? Therefore, AI doesn't attempt to replicate "stolen art," it learns according to it's database to generate the theoretical anatomy of what it's attempting to newly recreate. The Art that it uses is as grand as the internet itself. Not relying on any one image, but learning as we do when we draw what a house or a woman or a car looks like. That's why almost all AI images today come out as unique never before seen image like "Obama on a computer" or "Bob Ross but instead of Bob Ross the painter is outside, in an enchanted world, and the painting is on fire and also Bob Ross is a fish."
So, no
one's out here telling you guys AI is going to be replacing traditional artists, AI is just a different form of how an artist can create art. Just like photography, AI seems to be doing fine on it's own as a genre to recognize. AI artists can train their own models, make their own Loras, even make their own styles no one has ever seen before. The limitless reach of working with random compilations is only as much as a user can operate with the AI awaiting to work on a new image. The person begins an instruction, the AI responds, the person responds back, and the image is created by both agencies --this is how humans work with tools. Neither is anyone saying digital artists can't use AI to compliment your
own workflow. This blend is actually what I call the future. AI is an option, it's never a requirement or necessity as much as any other genre or convenience.
Thus, I can likewise ask the critics, can you trademark the color yellow? Can you trademark the shapes of houses? Can you trademark clouds or fields or flowers? No. But you can trademark your own paintings as a whole, complete Intellectual Property--though that doesn't mean you can sue anyone that uses your IP if what they create doesn't even reflect your original 100%. That's because AI images don't attempt to replicate any image 100% in pixel, color, or composition. Neither can you prove with evidence that an AI artist used someone's specific art piece among the millions of uncharted innumerable likelihoods. AI doesn't need to replace any genre, it is it's own genre.
--AI ART IS JUST 1'S AND 0'S --
Well, yeah. That's a captain obvious argument and dumb down response to how digital media works because media is transmitted by electricity.
This argument has even less weight these days when electricity is so common place, that a tv screen can display a picture of dog food all year long in modern shopping malls.
1's and 0's are the fundamentals that people aren't even educated enough to talk about. It's a cheap way to bring down decades of technological advancements to those early 1950's binaries.
The real topic people don't want to talk about is how we are all able to work in digital art fields and how society views technology in an ignorant and negative view.
Again, CathryneDelamort has made another great post about their actual workflow to enlighten people into how people use AI to create Art.
There was significant intentional, experimentation, and effort that went into bringing an idea from my head into an external medium. Many would say that this intentional translation of an idea to an external medium is the very definition of art.
...
People say that this new AI generated art is unlike anything that has come before and that it's just copying what's there. I will refer those to the invention of photography, Magritte's La Trahison des Images (roughly "The Treachery of Images" or "The Betrayal of Images"), and Duchamp's Fountain. These are but a few revolutionary techniques and ideas that revolutionized how we create and think about art. Generative AI is one such pivotal moment in history, and those denying its validity and refusing to explore its potential will inevitably be left in the dust.
(https://www.deviantart.com/cathrynedelamort/journal/It-s-not-art-because-the-AI-does-all-the-work-949716003)
--THE RISE OF AI HATE AS ITS OWN GENRE --
Why is it so difficult to accept that Art is what you create, and not where it's created?
Art is subjective, but an artist is a person. You can be an artist with any tool, even with computers.
As
a real life artist, that now works in digital media, I strongly believe
that it's not the tools that make the artist, but the creative skill in
how we use a tool. The biggest stress I have
as an artist is getting others to just be creative, to not think of
painting as a traditional form, or poetry as a couplet blueprint, or digital art as a lifeless pixel.
It's
about using your platform, your tools, your wits to be free and
expressive in your designs. That's what it means to be an artist. You don't need to be talented to make art,
you just have to be free. Being talented comes with practice.
That's why I love
seeing people using AI to make images no ones ever done before. I
get amazed at seeing all the different ways people have used AI to make a design. The "real" artists today are just a bunch of bully snobs whenever an AI piece is shared. It's really pathetic behavior. People hate any form of AI because they feel threatened or insulted when they should be supporting a new media of Art. If AI is so unique, I really don't understand why people are so against it. It's like if Photography came out today and people rush to each post saying "clown" or "fake art." The hatred against this media of Art is so petty, to watch one artist ridicule another new kid on the block. You don't see me going over to the photographers or anime illustrators calling their images "weeb" or "fake art." I personally don't like furry art, but I don't go harassing people that post those images with mockery and shame. But, it's okay to do so on AI images? Even when websites allow separate tags, AI users still get bullied by random groups of people that target hate their opinion. And for what? Ignorant rumors about how AI is done? Do people even understand how AI is not stolen? Or did most just go along with that narrative as an easy excuse to hate the genre? Is it easier to hate AI than it is trying to understand the work people put into it?
The elite snobbishness of some artists is a real problem in the art community. This is why I never posted on sites like Art Station, because I always felt like my drawings weren't good enough compared to AAA studios and other college graduates. I think, even if I had no hand in AI, and still only used digital drawing, I wouldn't join the bandwagon hate on the galleries that host AI images. I would just feel like I did before, awe struck about a different way to make art. Now with AI, I feel less welcomed because my art isn't just in AI, but in a variety of digital tools. I wouldn't be a digital artist if I wasn't also able to know my way around digital tools, would I? And yet, this whole AI drama has just proven that certain "artists" are real bullies with a huge ego of entitlement, as if making Art the traditional way was so rare, unique, and impossible to learn.
I find, that most hate is largely from amateur artists or hobbyists. Because actual college rigid artists wouldn't feel so threaten to a new genre of Art.
--Data sets --
So to those painters saying, "AI is using my art," the answer is, it doesn't work that way. AI uses images for training, not for replicating outputs. It uses pictures of dogs, for example, to learn how to structure the shape of a dog, but it doesn't attempt to output a singular image it was trained on. Rather, it would depend on multiple photos of a dog in order to output a new image of what "dog" is referring to. This is just a surface level example, if you really want to go into the morality debate, be prepared to hold onto your argument through the deeper and more complex ways AI is coded.
AI uses a massive dataset that doesn't cater to any "one" single image to reproduce. A simple example of this is with Kermit the Frog. If you're using AI checkpoints that are anime-based, it would be really out of the ordinary for that database to be trained using different pictures of Kermit the Frog. But, even if you are using a checkpoint that does realism, you might have an easier time recreating Kermit the Frog. Though, even then. It would be nearly impossible for any attempt to reproduce someone's exact reference. This is why most people depend on specifically trained Loras to do a proper Kermit.
So, in short. Yes in the same sense that google might also have a copy of your art that someone else shared. No, that people aren't attempting to, or are skilled enough to replicate a copy of your art pixel by pixel, form by form, and palette by palette without using special editors or Loras. And if someone tries, isn't that just fan art? Like using AI to make Spidermans, and supermans. If it's really close to the source reference, wouldn't that just be plagiarism risk anyway? More on this in the copyrights part...
--AI autonomy (bot) vs AI user (human)--
So, I think the most common issue today, for non-tech people, is trying to get out of the misconception of a Hollywood autonomous AI machine stereotype, and understand that its people whom are using AI tools. Not AI doing all the work.
Whether you're a human working on a canvas or a human on a computer, it all goes back to creativity that defines the artist.
Just give it a try. Go ahead. Open up photoshop, look at the blank canvas, and create something artistic. It's the same on any digital platform or program you use, not just in AI.
I would make a new argument that artists already have an advantage using technology, but it's the technology tools themselves that become the biggest learning curve. Just like someone that doesn't do any Art, the technology setup and tools might be simple, but creating Art will be the learning curve.
This is what I mean, art is about creativity first and then time with skill and practice from the novice to the experienced. So the same argument that a "real" artist tries to pull of by saying "it takes skill to do a real painting" is only telling half the truth because skill comes with practice and practice can be developed no matter what platform you choose to do art.
To
use pigments in order to do realistic paintings. To use cubism to do
new perspectives. To use digital composition to make new artworks, even
to use AI to fill in the gaps, edit a picture, or get you started on a composition piece or work with generations. It all comes down to the novice artist adapting into a skilled designer.
That's the most important thing people miss about AI Art, if the design was made by a human or a bot. If the designs are made by a human, then that means even digital media is safe with the amount of artists using photoshop, MSpaint, 3D modeling, and AI tools.
If the designs are not made by a human, then they would be similar to what a standard program would reproduce not matter the variation and sequence.
(Bob ross in the digital age)
The difference between a human designer using AI tools, and a simple bot automatically using AI to make an image, is that anytime a person is involved, there is an intelligence of creativity at work, the same as the intelligence stance that a physical artist uses. I would argue, this ability to be creative, and not just random in generation, is the key that defines when a person is using AI, and when AI is generating automatically.
Such as cases where an AI program can be told to execute a number of solo outputs, so too can it also be told to "improvise" throughout it's order to create random images without any refinement or structure. That computers may be able to even mimic how humans are creative, but that limit is too great for any bot to 1. create a random image like a creative person does and 2. be able to refine the creative parts that make a design reflect what a human would do.
Thus, there is really only 2 ways an AI program can be self automated. 1. To generate the most standard outputs through a long curated list of words, settings, images that a person would do. 2. To be specially programmed to also after refine the first step in order to produce some creative design a normal person would do. Lastly, a person would have to monitor the outputs of a thousand automatic images to find any that can make sense and rival something even remotely close to what a creative person would make.
Finally, after all this effort, a programmer, has to claim that they used an AI bot to make an image.
Vs
An AI user telling outright, from the start, that they were the ones that made the image.
So, in truth, the debate of a bot AI image vs an AI user all goes down to whether any user can claim responsibility to verify when an AI piece was done by them or not. This check is important when it comes to competitions and awards, because if an Image was done by a human rather than a bot, that human would be able to disclose the settings used as well as the prompt. Sort of like in a videogame competition, there has to be a live monitoring to catch when a person on a computer is involved and when they are using bots or cheats. In my own opinion, I wish people could just be honest and not use bots especially if you actually care about AI images. This is only because as a creative, I wouldn't like to showcase my own personal workflow and practice I have developed that reflect where I currently am for others to mimic. Just like a normal artist, I don't mind when people emulate my work, but I do mind when people plagiarize my originals.
--Copyrights--
This ALL goes back to the reason why so many don't want to be public about being an AI user. Because this again goes back to the same old western stance of copyright issues that many are being denied to and hated, simply for being AI users. When it comes to copyrights, because people assume ALL AI images were made by a bot, and not a person using AI, many copyrights fail to acknowledge that
1. AI is a tool being used to create a new original digital art designs 2. That its being made by a sourced author, like a person.
To not accept AI is the biggest flaw in our digital age where digital images don't have the same weight in court whether its made in photoshop, made on MSpaint, or simply digital drawn. We're all in the same boat if our works are getting hate targeted by people outside the digital art community for not using traditional physical tools.
This is the 21st century, I think its very clear there are numerous amounts of artists who are solely interested in working with digital tools like Photoshop, Krita, Digital Drawing, Mspaint, 3D modeling, and AI composition.
That's why I don't tell people when I do or don't use AI outside the AI community, because so many think AI is a computer made image design, instead of a human using a computer to make a new image design.
AI is so dumb-new that people don't like to find out when something was made with AI tools. Even further, that AI infringes upon replicated original art. Like that means every artist out there has a wild card to claim that they were infringed upon even though the AI user never used their art.
To get upset about AI is one thing
, but to have someone randomly accuse you is the dumbest thing that could happen today. Imagine some random dude took you to court because they think you stole their image. Its highly unlikely to happen. but it will bring up drama if it does. But this free to sue anyone that uses AI only highlights the difference between Art and Art Ips. Like between a mouse, and Mickey Mouse.
Like another user has said (Yowanas) " It has even gone so far that people who actually did draw things have gotten banned because their images looked "too good". Hopefully things will settle down and AI becomes just another tool."
Thus, I can defend myself equally that I make my own designs (whether using AI tools or not), but whoever sues me because of AI has to provide proof that any of my designs reflect that person's original art piece. It wont. That's because AI users aren't in the interest of replicating art, they're busy making new art.
--Final Thoughts --
I would highly suggest to go see how people actually do AI art, what it takes to learn, and how new people are different than experience people.
So again, new Art methods NEVER replace older methods, they only introduce a new way to make art. -- In
the future, people will grow to be more accepting of digital artists,
but for now, all we can do is our part, at the best we can, without
stepping on any toes or doing anything morally wrong that people can
accuse of theft, plagiarism, or just straight shoving AI into non Ai
categorys. In return, I hope people will begin to see the genre of AI
with a bit more respect, that AI is not this stereotypical hollywood
idea of a terminator or breathing entity that takes over the world --no
no, that is just movie magic, most AI is done with human calibrations . All that we have to do better is organize how
to reduce copyright infringements, do ethical training, and avoid
artists that want nothing to do with AI.
Especially on open source,
only YOU can restrict yourself of using Loras in right methods without
deliberately reproducing established IPs for commercial use.
Lastly,
digital art vs physical art are both means of creating art. It's been that way since people realized they can make art out of bit pixels and space invaders. The
uneducated attention in AI gives computers all the credit, but most digital images are worked on by people. Digital Artists are people working with digital tools like 3D, digital drawing, music DAWs, and so forth.
(Pokemon artists have long mastered the pixel art designs)
So the question should
never have been "was AI used?," but "who did this AI image," Because to ask if AI was used is as silly as asking if a digital imaged used photoshop or any editors. Most do, but there are already standards for serious events like competitions and awards.
Hopefully, people can be honest with
their fans, and disclosing their private workflow should only be towards lazy people suspected of using automatic bots or if AI prompts were simply copy-pasted or if someone thinks an AI image was entered in a non AI competition.
But, there are always a few bad apples and fake users.
