An upheaval is emerging in the art world. Artificial intelligence, once confined to the pages of a sci-fi paperback, is now part of our daily lives and challenging our view of creativity. As AI-generated art begins to grace galleries, newspapers and music studios, both artists and techies find themselves at a crossroads. AI confronts our perception of credit and originality while expanding the boundaries of what art is. In this brave new world, algorithms are the new artisans, creating a future where brushes of code become indecipherable from strokes of genius. The question is, can society adapt both ethically and culturally to harness these advancements while maintaining the integrity of human artistic expression?
In reaction to this burgeoning issue, the Springfield Area Arts Council's new program, Red Light Night, was started to "allow the public to discuss provocative and controversial topics within art," said executive director Bella Szabo. The first of these was held May 28, with the topic focused on AI in art and how to spot it. The well-attended meeting was moderated by three local artists of varying mediums, Cee Jones, Devin Larson and Bri Skeels.
The prominent ethical concern debated was plagiarism and how it's defined. Most AI art generators scrape the internet for as much information as possible to use in their datasets. When an AI artist types in texts prompts, the generator uses that stored dataset to create images from those prompts.
"One could argue that AI outputs would not exist if not for a human's creation," said Jones. Skeels posed the question: "Someone takes the style of an artist and uses that to create an image – does that artist own that style? Is that plagiarism?"
The fact is, artists have been inspired by other artists since the beginning of time, which previously hasn't been considered plagiarism. However, a class action lawsuit was filed by 4,700 artists in February. AI critics believe the problem isn't artist influence, but instead being put into a dataset without consent. The datasets of many of these AI-generators are black boxes.
"I don't want to sue because it looks like my art, I want to sue because my art was included in a dataset without my knowledge," Larson said (Larson is employed as a graphic designer for Illinois Times).
Another issue is the cost of this technological advancement, a "video killed the radio star" fear, similar to the uproar that accompanied the rise of photography. Portrait and landscape artists initially believed the camera would lead to their ruin, but the impressionist movement came along because photography liberated those same artists from the confines of mirroring the real world. Could AI have the same effect, but with an even more dramatic impact?
Attendee Nicholas Mulholland, who works as a graphic designer, shared his concern. "There are also inherent biases in coders and programmers; they're overwhelmingly white and male," he said. This could worsen the existing gender and racial inequities in art by not providing sufficient representation in the algorithm's datasets.
However, it is also possible that a machine might be better able to filter out bias than a human. By incorporating human oversight, fairness metrics and inclusive training into the AI development process, systems could be created that uphold justice without pesky human prejudice. Jones pondered, "If we can clean up the datasets, if it's all free to use, at that point, are we OK with it?"
The fusion of humans with AI-driven efficiency could allow artists to focus more on their craft or have time for better research. It's also a powerful mechanism for a more inclusive and accessible art world. Not everyone has the option of music lessons or the mobility to paint a canvas. With AI, creativity is easier to come by, allowing for more a more diverse group to join in on the artistic expression.
Unfortunately, with more participants and more powerful AI programs now capable of producing masterful art, more artists, writers and musicians might soon find themselves out of a job. Case in point, this whole article was written by AI.
Just kidding.
But do you think you could tell the difference, and more importantly, would you care?