Vincent Matthew Franco
Social Media Editor
Artificially generated art, or A.I. art, has the ability to bring to life the wildest concepts on a computer screen. Ask for a rendition of Edvard Munch’s “The Scream” in the style of a Wes Anderson movie, and you will get exactly that. The possibilities are endless, as well as the backlash from traditional artists.
Over the last few months, A.I. image generator websites like Midjourney, DALL·E 2 and many more have grown in popularity for their seemingly effortless way of conjuring up the most ridiculous prompts into crisp illustrations. Updates to these systems over time will allow for even more advanced results. The rise in popularity has brought with it controversy and legal actions from both sides of the A.I. Art debate. At its center are the concepts of originality and creativity and what makes art “art,” which are also the basis of Intellectual Property lawsuits that have sprung up around A.I.-generated art.
For A.I. artists like Cameron Early of Michigan, this medium can be used as a new way to make art or even just as a tool to help artists excel in their ways of creating.
“It’s just one of those things where you can’t stop it, it’s going to keep getting better, more people are going to use it,” Early said. “It’s going to be used for advertisement. It’s going to be used for everything. It’s going to take over everything creatively, essentially, and just get on board or don’t.”
These text-to-image A.I. Generator websites, like Midjourney, work by simply typing a few words into a text box and the generator creates an image based on your prompt. The more descriptive the prompt, the more accurate the image will be. Early has given up whole days trying to come up with the right prompts in order to make the picture in his head exactly how he wants it on screen.
Making A.I. art under the handle “mrmonstersauce” on Instagram since October 2022, Early’s attraction to A.I. generated art began when he realized he could create not only anything and everything but all in his favored aesthetic. He takes inspiration from his massive collection of monster toys and the vintage Halloween collectible items that flood his room.
As you scroll down his Instagram page with over 700 illustrations, those inspirations are all evident. Each of his illustrations portrays different ghouls, goblins, and swamp creatures in hyper-realistic pulp-like images in settings ranging from the 1960s to the 1970s. They are not just creatures standing around either. They are all placed in locations and situations that look like they were taken straight out of an R.L. Stine Goosebumps novel.
“There’s a lot of ways to approach it for sure,” Early said. “I sit around and think of weird, creepy things in my head that I want to see. I punch it to the best of my ability, sometimes I don’t get it right the first time and then I’ll change a few words and I’ll try it again. I’ll just keep rolling until I get an image that (I think is) the one.”
But how much input does someone like Early actually have regarding the final image?
“Realistically, not that much,” he said. “You type in all these words and things you want to see, but the aesthetic that it gives you and the real realism that it gives you has nothing to do with you, absolutely at all.”
This is where it starts to get controversial.
Like the many artistic and technological advances before it, A.I. presents ethical as well as legal questions. The most concerning problem is its propensity to take the concepts and sometimes blatant parts of images from artists’ online portfolios without permission.
Los Angeles lawyer Matthew Butterick is a writer, graphic designer and computer programmer who is co-counsel in two lawsuits involving A.I. generators and copyright infringement issues.
“We’re taking another step toward making A.I. fair & ethical for everyone,” Butterick wrote in one of his blog posts.
One of the cases is a class action lawsuit brought by three artists and illustrators, Sarah Andersen, Kelly McKernan and Karla Ortiz, who are suing the A.I. generators Stability A.I., MidJourney and DeviantArt for copyright violations. In their lawsuit, the artists allege that these A.I. generators used their work without their permission or compensation to build training data sets.
Training is when images and the captions to those images are taken from the world wide web and used as a reference, essentially turning them into digital noise and from that digital noise, a new image is created. Even though the image looks original, these lawsuits claim that what it really is, is a collage of a billion other pictures and illustrations from millions of other artists.
The suit also claims that copies of the training images were being used to generate A.I. artworks that are in the style of particular artists, thus impacting the market for their original work.
A separate class-action suit brought against Microsoft, its subsidiary Github, and Open A.I. also alleges that the companies are engaged in copyright violations and piracy. The suit claims that these companies scraped copyrighted material from publicly accessed open-source sites on the internet to train their A.I. systems Copilot, a Github product, and Codex, an OpenAI product. But they removed the licensing information from the codes, which would require the artists to be credited for their work.
“There are pro-A.I. people who feel that if you put your art on the web, you’re implicitly saying that it’s okay to be used for A.I. training,” Butterick said. “I think the question right now is, legally speaking, what should be the limits of how much A.I. training is okay?”
“Are we in fifth grade?” artist Jason M. Allen asked facetiously. “It’s not hard to figure out. If you want to protect your work, then you got to take better steps than just plastering it on social media.”
Allen’s foray into A.I. art won him a first-place ribbon in the digital art category at the Colorado State Fair in September 2022 for a piece he made on Midjourney called ‘Theatre D’opera Spatial.’ The work is a prime example of what A.I. art can create and its effectiveness in merging different worlds of art together.
In the image, there are three women facing toward the back of a castle dressed in Victorian dresses inside of a Victorian-era castle. In the middle of the castle is a bright white vortex that seems to lead into another dimension full of futuristic buildings.
He defends his work against anyone who says A.I. Art is cheating and insists that it involves the artist’s own creativity.
“It’s just obvious to anyone that has a brain that the A.I. isn’t waking up one morning and saying, ‘Hey, Jason, I had this great idea… I made some images for you, you want to choose some of these and take credit for it, because I made them for you,’ that’s not how it works.” Allen said.
He argues that creating A.I. art is more than just pressing a button and goes beyond a word salad. To him, it involves a development process with a creative force behind it.
Dan Gentile is the culture editor for SFGATE in San Francisco and thinks that using A.I. can limit an artist’s creativity. Gentile covers all things art and can see how A.I. could benefit an artist in some ways but detrimental in others.
“I think that if in the right hands, this technology can be really a way to unlock creativity and create material that can be used to create broader pieces of art, but just relying solely on it is dangerous ground where it’s really easy for bullshit artists to pass out a few prompts that they’ve entered as a work of art that with a $1,000 price tag in a gallery,” Gentile said.
While the debate continues over whether A.I. art generators are illegally using the work of traditional artists, artists using A.I. to generate art may also find it difficult to protect their work.
Typing a prompt into an A.I. generator may not be enough to convince the U.S. Copyright Office of its originality. In order for an image created through A.I. to be deemed an original work, the creator has to show how much creative effort they put into it, proving that they have inserted their own creative touches in the artwork in order for it to be protected by copyright law.
While A.I. generated art is accused of stealing from the work of traditional artists, those working in creative fields, such as graphic design, may have something to worry about as well. Anthony Acock is the Chair of the Department of Art, Art History and Visual Communication Design at Cal Poly Pomona, and in his 30 years of experience, he has seen plenty of threats come and go to the graphic design industry. For him, learning to adapt is key to surviving these threats. One way of doing that is by keeping the human and cultural connection between the artist, the client and the target audience. The result of this will be a successful design, not like one of the million very slick-looking graphics coming out of Southeast Asian sweatshops, says Acock.
“Chat GPT or computer A.I.-generated art doesn’t do that,” he said. “It’s very slick looking. It will immediately provide, sort of instant gratification because it’s awesome looking, but it’s lacking the communication content that a relationship will have with the client and the audience and the designer.”
He is taking a wait-and-see approach to the future of A.I. and how it will affect this specific field.
“There’s a sympathetic fear if you’re trying to make a living as an illustrator, right?” said Acock. “That’s already a next-to-impossible nut to crack, it’s very hard to make it purely as a fine artist.”
Danielle Eubank is another artist who has been practicing for over 30 years as not only a painter but a graphic designer and is a former adjunct professor at the University of La Verne. She compares what is happening right now with A.I. to the introduction of photography in the 1800s.
“Did it change other kinds of art? Yes, absolutely,” she said. “And that’s what led to abstract art because people didn’t feel as compelled to replicate things in a quote, unquote, photographic way, it freed them up.”
Just like A.I. art is viewed now, photography was also seen as a way to create what was thought to be a cop-out to “real art.” The evolution of cameras and how they are used has gone far past that. Because of the new pathways, it created to composing not only abstract images but other forms of art. It became an acceptable tool for creating art. This is why Eubanks says she feels no immediate threat to the rise in A.I. art.
“It’s another tool to help you out,” Eubank said. “It’s going to take a couple years for the dust to settle, and what it grows into might be something that we don’t expect, that’s pretty common too.”
With the future of A.I. art still not set in stone, there are no signs of it slowing down any time soon. Midjourney has recently announced a version 5 update to the renowned text-to-image software, only six months after its V4 update in December of 2022.
This V5 update produces more photorealistic images and comes with more features for users to play around with. In a tweet announcing the update, Midjourney claimed that images will now be “more coherent, sharp, and beautiful.” The tweet also says that they’ve added a “raw” mode that lowers the “opinionatedness” of the software and will allow for more creative decisions to be made by the user.
A.I. art has gone further than just creating images, it has also made its way into the creation of videos. People can now create videos just as easily with a simple text prompt, identical to the A.I. generators formulating illustrations. Identical to the rough early versions of A.I. image generators, the videos produced by the A.I. video generator websites can be comical but also disturbing at times to see. The results from sites often show warped faces and distorted human figures.
The Republican National Committee has already made A.I. art and video technology a tool in their arsenal for smear campaigning. Released on the GOP’s official YouTube channel, a video created with images made on Midjounry and DALL-E depicts what life in the U.S. will look like after President Joe Biden is re-elected President in 2024. The slideshow depicts a fentanyl-ridden San Francisco, foreclosed banks, and hordes of immigrants flooding the borders. This is far-fetched from the funny and comical images that people may be using A.I. generators for now, as it shows the potential role A.I. technology has in promoting fake information.
The arguments swirling around A.I. art as well as everything else surrounding it, are going to continue. But one thing that is going to change for certain is the way art is looked at by people and the laws that govern copyrighting creative work. Artists like Jason Allen will help bring new points of view to the ongoing debate over what art is and what it means to be an artist. While lawsuits like the ones Butterick is involved with will help define the reaches of the law when it comes to protecting artists and their work.
Vincent Matthew Franco can be reached at firstname.lastname@example.org.
Vincent Matthew Franco is a senior journalism major with a concentration in print and online journalism. He has been involved in journalism and print media in high school, community college and is now at the social media editor of the Campus Times and a staff photographer for the Campus Times and La Verne Magazine. He previously served as arts editor.