Degeneration in the generation of A.I. art

What A.I. art thievery may mean for creatives now and in the future

Tara Winstead / Courtesy Pexels.com

Where ChatGPT threatens to displace writers in their craft, A.I. (artificial intelligence) has been clawing after artists’ works for years now. Many have had enough. In a New Yorker piece released earlier this year, a group of artists — Kelly McKernan, Sarah Andersen, and Karla Oritz — claimed that there is a fight to be had, evidenced by their class-action lawsuit against A.I. imagery generators Midjourney, Stable Diffusion, and Dream Up.

Truly nefarious or just unexpected, A.I. art poses a host of ethical questions about its use. Art thievery is one of them, as stated before. A.I. imagery generation necessitates an algorithm to “learn” a specific aesthetic or artistic genre by analyzing thousands upon thousands of images, according to American Scientist. Creatives fear that their work will be taken from their sites and filtered into these algorithms, their labor becoming a pawn in the crude generation of shallow images.

Often, A.I. “artists” argue that their process of appropriation is not novel. Roman sculptors fashioned their pieces based around the ideas of ancient Greeks. Even now, sampling songs is deemed as an art form, where decades old tunes can be reimagined to fit modern tastes. But A.I. art is decidedly different for reasons of scale and human involvement.

For one, A.I. artwork generation and creation of aesthetics depends on hundreds of thousands of images and the hours of labor which go along with it. Where one artist may take inspiration from the work of another, the infringement of A.I. is almost infinite in comparison. And since users from all over the world can use A.I. image generators with ease, tracking which works have been stolen from who gets more and more arduous.

In a similar vein, A.I. art is merely a regurgitation — a rudimentary blending — of styles that creatives may have spent decades perfecting and defining. The process of A.I. art rejects soul at its essence; it contains no remolding or adaptation, no real new thought, no human depth and originality. Where people will create regardless of each other, A.I. art’s existence relies on humanity to survive. In each rapidly produced reimagining of a realistic Bart Simpson or a pair of futuristic twins, there can never be the heart of an artist. There will only be the curiosity and desires of an experimenter.

Perhaps the most pernicious aspect of A.I. art is its grotesque extension of female objectification, often to lengths that are disturbing to gaze upon. One scroll through the #aiart hashtag says enough. In between images of magical beasts and renaissance-esque cartoons are depictions of women in the uncanny valley with body proportions aimed to ignite sexual desires.

Just as fanservice in anime favors enormous breasts over female intelligence and character arcs, A.I. art demeans womanhood by reifying the same sexist ideals permeating the digital world in video games, social media, and T.V. Women exist only for pleasure, and A.I. art perpetuates this idea — now, at a much, much wider scale.

It is vital here to emphasize how such technologies are not created in a vacuum. It is people who create the algorithms A.I. art is based on. It is people who choose the thousands of images that refine the algorithm and help it “learn” a particular aesthetic. And it is people who engage in art thievery, who create unrecognizable visions of the “perfect” woman, and who use these digital tools to reify sexism instead of finding avenues for empowerment. Yes, there is an endless list of programmers, designers, and consumers who may continue to degrade the creative space if nothing changes.

This is not to say that misogyny and art have not been intertwined for some time, well before the advent of A.I. technologies. Great artists like Georgia O’keeffe and Lauren Greenfield had to fight for their place in the male dominated world of artistry, thus amplifying female voices over sexist ideations. This is also not to say that such individuals will fall out of favor with the public and will lose their ability to influence others with their works. However, with A.I. art — through its blending of works and its perpetuation of sexism — significant, feminist pieces may get lost in the background in the search for the feminine “ideal”. Audiences may turn their eyes to instant gratification, like the overexaggerated female form, rather than human-driven art.

Since it is people who are producing such ethical conundrums in need of analysis, it must be the same people who paint a brighter picture of the future. After all, it doesn’t seem that A.I. art is going anywhere anytime soon. A.I. advocates must listen closely to the struggles of current creatives, so the market may be habitable for both parties. They must introduce some restrictions on the kind of pieces that can be produced, or at least take a firm stance against exploitative, misogynistic art.

The path ahead looks bleak. Artists are still vying to be heard by the general public and the image generators taking advantage of their hard work. Regardless of what the outcome may be, it is imperative that A.I. artists keep real creators in mind when trying to make A.I. art mainstream. After all, art is nothing without the hard work of people expressing themselves and helping others do the same.

Mary Kurbanov

Mary Kurbanov (she/her) is a writer, editor, designer, and recently graduated Media Studies major at the University of Virginia. Mary has served as Editor-in-Chief of the V Magazineat U.Va., the Newsletter Editor and Arts and Entertainment Senior Writer for the esteemed campus news outlet The Cavalier Daily, and graphic designer for film magazine Page 2 Frame. When she is not dedicating herself to her craft, Mary likes listening to music, enjoying a good meal with friends, and traveling whenever she can.

https://www.marykurbanov.com/about
Previous
Previous

In urgent defense of Sofia Coppola and her upcoming film "Priscilla"

Next
Next

La petite anglai