All Ai is built on the programming of images in existence.
Most Ai uses a process called “web scraping” to train it’s technology. (TowardDataScience, 2022)
If we think about the subject of a cat.
There are 6.5 billion of pictures of cats on the internet.
These images are scraped and programmed into Ai
Imagine every breed of cat in existence and every artistic rendition, painting, sculpture, 3-d character, illustration and photo of a cat is contributing to this knowledge in Ai of what “cat is” in most ai generators.
And then can be called upon with other identifying parameters.
This is then the same process for any subject in existence that identifies environment, art medium, lighting, styles, art movements and pretty much every artist, photographer, art piece and art movement in history of pretty much every single thing on the internet.
This is like an entire “collective consciousness” of the human experience of objects and things — how they look, what they are. Very similar to how the internet is like a collective consciousness that we can call on for research and when those parameters are mixed a “wholly new work is created”, not a filter slapped on. But the ai uses it’s knowledge to create something new!
This is what makes most ai generators so it so powerful because it gives on many ai generators you something you can use commercially…
Why? Because of how different it is.
As long as it’s not a trademarked character (ig. Mickey) or a celebrity likeness.
Those characters and likeness are protected under the law, but an artists STYLE is not.
According to copyright. gov:
“Copyright does not protect: ideas, concepts, systems or methods of doing something.”
You may express your ideas in writing or drawings and claim copyright in your description, but be aware that copyright will not protect the idea itself as revealed in your written or artistic work.”
Note: You can’t copyright the subject, you can’t copyright the METHOD <– which where style / medium fall into.
And thank god for this.
It would cripple ALL creatives of there was only one style permitted per artist. We would only have a handful of artists that could legally create art.
Artists say about Ai generated art “My art is stolen”
If we’re gonna be fair here are you talking about style? Cause style can’t be copyrighted..
And to be fair, “The whole internet is stolen” to create these machines.
Here’s how they get away with it legally —
There is one tiny caveat to how they can do this– This is the secret to the whole thing!
Are you ready for this?
The production of the pieces by the ai are generally “Not truly a derivative” to a copyrighted work.
BOOM!!!
Meaning Ai found a way to …….Just change it enough to have a new perspective…. become a “new work” under the law.
It’s not a trace. It learned from its 500 billion image knowledge of what something is,
Lemme repeat for those in the back: it’s not a trace or lines from the original!
And that “new work” is skirting the infringement, why?
Because you can’t copyright style.
Style is why in the cubism movement so many paintings looked similar.
Style is why there can be multiple comic book or manga artists that have the same look and feel, but aren’t considered ripping each other off, but instead are a part of a specific industry.
Style gives you freedom to learn how take what’s currently in existence and REMIX it to something the world hasn’t seen… but yet feels familiar.
Sidenote: do you really think these big businesses be so stupid to build an entire business on something that would get them shut down based on basic copyright infringment?
They’re not that stupid.
That’s why all these ai sites aren’t getting shut down for copyright infringement left and right because under the law you cannot copyright “style”
And THANK GOD we don’t have a copyright on “style”… or literally the first artist of any art movement would have been the ONLY one to use that style and the world would not have art movements!
The Ai can create works that look “similar” but not exact. Just like ANY artist can do the same thing. <– which is also frowned upon.
If I love Andy Worhol and I want to do a style of portrait of a “Boston Terrier” in the style of Andy Worhol — no one can say.
“Hey only Andy can do that Style … now I’m going to sue you!”
Here’s what copyright.gov says about “Derivative Works”:
The following are examples of the many different types of derivative works:
• A motion picture based on a play or novel
• A translation of an novel written in English into another language
• A revision of a previously published book
• A sculpture based on a drawing
• A drawing based on a photograph
• A lithograph based on a painting
• A drama about John Doe based on the letters and journal entries of John Doe
• A musical arrangement of a preexisting musical work
• A new version of an existing computer program
• An adaptation of a dramatic work
• A revision of a website
NOTICE HOW IT DOESN’T SAY ANYTHING ABOUT STYLE.
Here’s what you need to understand about this:
Don’t do this!
if you’re using prompts that are specifically calling out a SPECIFIC PIECE of ART in your ai prompt… you could be accidentally creating a derivative work.
And/or referencing only ONE artist (especially if they’re live and working) DO NOT DO THIS. Just STOP referencing working artists.
Do This
There is a caveat to this: Parody. You can create a political commentary or joke about a trademarked character to prove a point.
If I wanted to write a comic and use let’s say “The Joker” to make a point about turning into a villain. I could do that!
Or you can make a joke about anyone or anything for any reason.
If you feel like Disney is destroying the world, you can make a Mickey made of nuclear warheads and now it’s gone from “art for profit” to “political commentary” which falls under freedom of speech and Fair Use laws. And yes this can still be sold at a profit but it falls under that law.
Basically a Mickey warhead art piece is not competing with Disney.
Don’t do this
Is it a good idea to do this just blatantly COMPETE with an artist ——without any shift or change of perspective?
NO.
…Using another working artists style can create confusion in the marketplace
…It can be proven they’re losing business
…and the truth is there are unethical people who are producing work in their style and taking their jobs.
And that SUCKS.
. Do This
It’s important to note here, the vast majority of people who are using it are not doing this.
The vast majority of people I know using it — are using it for personal projects, generating images of their imagination, illustrating meditations, project ideation, remixing styles to something unrecognizable.
They aren’t slapping together a portfolio with one artists style and being like “HIRE ME!
“I can make art just like Alex Grey!”
They’re “remixing” and creating their experience and perspective into the Ai generated art.
And if they aren’t — they should be.
Fuck the thieves that are.
And let’s face it — if something is “too similar” in the eyes of a judge — really the judgements can go either way!
Example Andy Worhol lost a legal case of use of an image of kittens in one of his paintings, he found at an airport.
The Worhol estate won for a case regarding a misused image of the pop star Prince that fell under “just changed enough” to not pay the original photographer even though usage was outside of their usage agreement.
Same issue “this is a derivative work of a photographer” and two diff legal outcomes.
If a specific artist is used and that artist can show “lost wages” they may have a copyright infringement through derivative case.
Again, don’t build anything off the style of one artist.
Now there are BLATANT ways you can manipulate the system and get yourself in trouble!
An UNETHICAL approach in a prompt reference — here’s an example
Don’t do this
” Create x in ONE working artist in a prompt and/or the exact name of a piece” or saying “in the style of x artist”
Do this instead
Create prompts stacked with a variety of modalities, subjects, and various styles, and if you do reference artists — reference many — not just one.
Remember JUST ONE is Plagiarism — Many is RESEARCH.
If you don’t follow this guide — it will likely get you a TON of online hate as seen by Joe Rogans post where he referenced an Ai generated piece that gave credit to “Alex Gray” as the prompt was used to produce the work…
SideNote: if you actually look at the ai generated piece it’s “similar” looks like a “knock off” but still isn’t an exact lines used replica of any of Alex’s pieces — this is still NOT OKAY. Because he is a working artist.
Dont do this!
There are some threads going viral right now controversial back lash of artists showing “This is Ai” and “This is the Original” with subtle differences.
Saying “See Ai Steals!”
These examples were actually NOT ETHICAL ways to argue this!
The person sharing this had uploaded an image into the system, referenced the artists name, and piece and tried to recreate it where it showed very close output…
The Ai output was different but way closer than the original image.
If you really want to steal this way you could — but would be like putting into the ai to blatantly steal!
Here’s what they’re prompt likely looked like
“Create the Mona Lisa, Here’s a link to a mona lisa picture, make this image of the mona lisa by Leonardo Davinci”
And then being like shouting “The EFFING AI Stole the Mona Lisa”
It’s like “no shit” because you stole it when you personally upload the image, told the AI to copy it, and called upon the knowledge of that artist to recreate it.
— in this case the USER programmed this to steal to make a point.
The same way they could have gone to the internet and stole it or painted something so close it used the same lines.
Do this
Since this tech is not going anywhere I have some ways for you to think about this
There are ways to “Ethically” use the knowledge of 500 Billion questionably sourced images. And there are ways to steal on them.
Don’t be an asshole.
Since Ai creates something WHOLLY different than it’s original inspiration, it can be MANIPULATED into creating something that is “TOO SIMILAR” — this is where the legal ramifications will happen.
I’ll give you a simple image for understanding, using “Klimt’s The Kiss” as an example.
If I go and put create “Klimt’s the Kiss” in an ai generator (and a REALLY GOOD Generator) this is what it will create:
See Image above: This is very similar, similar in style to the original. If I was to pass this off as my whole idea — that would be wrong.
Example “Hey guys look at this cool digital painting I made of me and my boyfriend Kissing!
Oh la la” <– not ethical — I didn’t reference the artist (even though the artist is dead).
If I was going to try to sell this commercially — I technically could because it’s in the public domain — and so could anyone else.
Ethically I’d still reference the artist to honor them! “The Modern Kiss Inspired by the Original Kiss by Klimt”
It’s public domain you can do that.
For a working artist now. Don’t do that.
And since it was also made by ai…”who owns the copyright?” use would be questioned unless it’s morphed and changed by a human.
What I could do or anyone could do is use this image regardless of copyright status because this version of the image is generated by a ROBOT….
…and robots can’t be granted copyright, nor can the ai creators.
The good news is — you also can’t be sued by a robot. Which is why you can use things like this commercially!
What you can be is sued by a working artist.
Any “is this a derivative ?” case could go either way, because the laws aren’t super clear!
….but if it shows that a case where an artist is losing revenue because of an artist creating pieces that are indistinguishable from the original artist…. they very well may have a case.
So, Here are The Rules That I Recommend Following :
1. Don’t use a working/living artists name in your prompts, unless you use 4 or more together.
2. Do not reference any working artists pieces by name.
3. Do not say “in the style of artist” UNLESS you are stacking many artists to create something WHOLLY NEW and UNIDENTIFIABLE by the original artist.
4. If you’re going to do a mashup of something as a creative exploration — Ie. Klimt and Darth Vader, or Wes Anderson directs Starwars — Realize these shouldn’t be used commercially. And reference the artists in your posts.
5. And if people bitch at you because they think you should HAVE to take the road of going to art school to create ai images in the painting style — give them a hug.
This is a really tough time for artists right now who have devoted their time to the craft and to feel it ripped out from under them.
(I also personally believe all of this is going to make traditional art MORE valuable! Fine art values will go way up. Because less people will be learning how to do it!)
Also realize that EVERY HUMAN on earth has a right to creatively express what is in their mind. And ai makes this way easier.
Especially for people who have mental health disorders.
One of the founders of one of the most popular ai sites, had reported that the main use case he’s seen for people using it is art therapy and personal projects.
I find this to be very true!
I have students who are illustrating a book for their baby to tell them about their birth.
I have a woman who just created an amazing deck of oracle cards of Black Mermaids.
I personally have created images of me being pregnant, that were based on a vision I had in a meditation.
This idea that only artistic expression is saved for those that are classically trained or naturally talented is frankly bullshit.
Most of the members in my membership ARE traditionally trained artists who are being smart and SHIFTING this technology into their work flow.
This is a shift or DIE moment for artists and creatives.
This is going to happen for ALL of us writers, animators, movie makers.
Shit there is even Ai that pumps out ENTIRE courses — you give it an idea and it makes the script, voice over, and video.
Which in theory could put me out of business, but I’m not afraid.
Why?
Because this tech is also going to put the capability of you to create your very own Pixar movie in a matter of hours in the next few months.
There is OPPORTUNITY for new creative expressions everywhere.
Being able to creatively express in ways you never have before can help you experience a JOY you’ve never had before.
This same debate has happened throughout history.
New tech comes out — people lose jobs — but they find new professions.
Suddenly the guy who works at Walmart with no education, can create the next mind blowing Ai movie that changes the world.
I see the vision is the world becomes a world of creators and beautiful things are EVERYWHERE.
And then the funny part is that imperfect things, real things will have more value.
My recommendation for Traditional Artists:
just realize this tech here to stay —- learn it! Use it for collaboration in your work!
create art ethically within the AI space,
encourage others to do the same — not shame them for creative explorations.
And ai generators:
be ethical
give credit
donate and or buy work to living artists you are inspired by!
Let’s hold people who blatantly steal accountable.
Don’t try to pass of AI art as your own.
State “created with ai”
Ai generated art needs to be its own league, not even compared to traditional.
And let’s create beautiful things together!
The good news the desire for even needing to referencing current artists is shifting.
The ai is getting so good, you don’t even need to anymore.
Jenna Soard have been in the art and design industry for more than a decade already and she have been in the forefront of this A.I. revolution from the very beginning. She wants to make the A.I. art world more ethical and at the same time accessible to all. Here she beautifully explains how A.I. art generators works and legal aspects of those. Pay her a visit to https://creativitythatconverts.com/ to see her awesome works and no, this is not a sponsored post.
Happy generating.