Written by Tate Moyer:
Introduction
“Once we lose this capability of telling what’s real and what’s fake, everything will suddenly become fake because you lose confidence of anything and everything.” These were the warning words of Wael Abd-Almageed, a professor of Electrical and Computer Engineering at the University of Southern California, during an interview with the Associated Press News this past week. His warning was in reference to artificial intelligence (AI) and the expanding role that these technologies are beginning to play in everyday life.
AI on the Rise
In the past decade, AI has emerged as a “mainstream technology,” increasingly being trained to execute a growing variety of tasks and functions. Jobs involving assembly-line style labor, information gathering and analytics, and telemarketing and communications are progressively being performed by artificial intelligence. In fact, Forbes estimates that an estimated 85 million additional jobs will be replaced by these technologies by 2025.
With the majority of these jobs, the replacement of human labor with these technologies appears logical. The use of machines to execute repetitive tasks and operations offers the possibility of streamlined production and increased proficiency. It could create a world where people have to do less tedious work and have more time for themselves. A circumstance that was largely unanticipated, however, is that these AI technologies have also started to penetrate the realm of creative and artistic disciplines.
Artistic Intelligence?
Indeed, disciplines that were previously thought of as uniquely human, including music composition, creative writing, and artwork, can now be generated by artificial intelligence. Softwares such as DALL-E, Midjourney and Stable Diffusion all possess the capability to generate digitized artwork in a matter of seconds. The debut of these technologies have not only sparked concern among the artistic community, but they have also highlighted an ongoing philosophical debate: if AI is supposedly not capable of original thought, then how could it be capable of creating truly original artistic pieces?
This is the driving question behind two new lawsuits that emerged last month, both filing legal claims against image-generating AI software companies for allegedly copying and processing millions of copyright-protected images without proper licensing. Frustrated at the prospect of their art being used for “inspiration purposes” without proper crediting, artists such as Erin Hanson maintain that “The original artist needs to be acknowledged in some way, or compensated.” Hanson, an impressionist painter, is only one of the many artists whose work has been widely replicated by open-source AI image generators.
Disputes surrounding copyright infringement of artificial intelligence have not been uncommon over the past decade. As AI and machine learning models continue to draw upon growing amounts of preexisting art, music, and code to create ‘one-of-a-kind’ pieces, the boundaries between original and reproductive works have become increasingly blurred. The line between inspiration and legally actionable theft is very, very fine, especially when the laws governing copyright weren’t written with AI in mind.
Legal Buffering
While lawsuits surrounding these situations continue to arise, society is left to grapple with the legal fallout that often occurs as a result of the rapidity of these emerging technologies. Technology grows exponentially, but societal infrastructure grows at a much more linear pace. Accordingly, the legal landscape often finds itself painfully unprepared to regulate for new technological developments. And indeed, we have entered into the age of the “Legislation Lag.” As legal regulations race to keep up with the ever-expanding world of the internet, we can only hope that there is some way to balance the immense potential of these AIs and the rights of artists.