The Advancement of Artificial Intelligence (AI) and Art: Ethical Issues and Legitimacy of AI-generated Art

I have been an aspiring artist since I was only a few years old. Even though I am not great at it, I have put many hours into bettering my art and trying to meet industry standards. I have always worried about individuals where art comes easily to them, but haven’t let that get to me. A few years ago another worry arose, that being AI-generated art. Not only have most of the art pieces I have seen by AI-generated art been beautiful, but I have seen the use of it within the industry of graphics design and logo design, and how clients/stakeholders are opting out of the ‘human touch’ for a cheaper option. With the world increasingly advancing with technology, AI seems to only be getting more and more integrated with our online lives (i.e. Adobe Photoshop AI-generative fill, Chat GPT, etc.) multiple questions have risen up within the past few years about the legitimacy of AI-generated art; is AI-generated art ethical who gets the credit of AI-generated art, should collaboration between Human and AI-generated art be allowed for award-winning, etc.

Zivy Epstein, a post-doctorate from MIT Media Lab, research focusing on the design and social sciences with regards to generative AI at Stanford Institute of Human-Centered AI, along with his other research partners has gone into who gets credit when it comes to AI-generative art (Epstein, 2023). Their research started with an ethical situation starring the AI-generated portrait called Edmond de Belamy (Figure 1) that was sold at a world-renowned auction house called Christie’s. Christie’s was founded by James Christie in 1766, which is known as “a world-leading art and luxury business” in the form of an auction house that auctions all sorts of luxury goods in over 46 countries (Christie’s, 2023). When Christie’s first marketed Edmond de Belamy in their auction show, it was to be known as “the first portrait generated by an algorithm to come up for auction”, expecting a price no higher than 10K for it (Christie’s, 2018). Even though this wasn’t completely false, there were many who had a huge involvement behind the creation of this ‘AI-generated’ art piece, who were not credited nor given even the tiniest sliver of the 432K, such as the researchers who aided in advancing/improving the program or even the Robbie Barrat, the 19-year-old who wrote and programmed the algorithm/code to produce Renaissance-style images with the program.

Some would argue that AI-generated art is not even that good, some with the opposing opinion (mostly those in the industry, trying to produce art with as little money invested into resources). This leads to another research study done by Jimpei Hitsuwari and his colleagues. Jimpei Hitsuwari is a current student aspiring for his doctorate at Kyoto University, researching psychology in regard to the aesthetics of haiku poetry. Hitsuwari and his colleagues did a study with the help of 385 participants who compared haikus containing 17 syllables that had 3 different variations of haiku poems (some made solely by humans, some by AI-generative power, and some AI-generated with human involvement). These 385 participants would rate the haiku poems on many different aspects (i.e., beauty, awe, image, etc.), and then identify which poems were AI-generated and which were made by humans.

In Figure 2, the average beauty score of human haiku poems was 4.15, the average score of the AI-generated haiku poems without the help of humans was 4.15, and the average score of the AI-generated haiku poems with the help of humans was 4.56. The results of this study done by Hitsuwari et al. clearly showed that human-AI collaboration when it comes to art yields a higher satisfaction score with regard to creativity and beauty. The assumption is that human-AI collaborative art would yield worse results than if it were just a human’s authentic art, yet studies suggest otherwise.

Hitsuwari et al.’s point that human-AI collaboration is more beneficial than harmful in the art world is further justified by artists that Kevin Roose, a journalist for The New York Times, with his article “A.I.-Generated Art Is Already Transforming Creative Work” published in 2022. A lot of artists, ranging from interior design to concept artists, that were interviewed had only expressed that AI-generated art/tools helped with enhancing their artwork, allowing them to add so much detail without having to fret and focus on the main concept. For example, Patrick Clair, a filmmaker who has worked on films like Westworld, used AI-generated art to help enhance the quality of his presentation to a film studio (Roose, 2022). This is just one of many examples of professional artists not worrying about the growth of AI-generative art in the industry and using it to enhance their artwork. As Aaron Hertzmann, a principal scientist and researcher at Adobe said “generative-AI technologies are yet another artistic tool”, for example, people use Adobe Illustrator to enhance their art and create more interesting art (Epstein et al., 2020). Nevertheless, these artists pay for it, therefore creators of the algorithm are those who enhanced the tool and should be paid.

With AI-generated art on the rise and the industry continuing to profit off of the art, I don’t think artists should be worried about losing their jobs, I believe the reason AI-generated art is seen as a growing problem for artists all around the world because of the lack of credit given to artists by industrial and large-scale use, therefore I believe any art that has used any sort of AI-generative characteristics, should credit researchers and creators of the algorithm that has taken part in the art itself. Previous research has shown that the collaboration between AI and humans will result in masterpieces that would be enjoyed by all sorts of crowds.

References:

Christie’s. (2023). About us: The history of Christie’s: Christie’s. About Us: The History of Christie’s | Christie’s. https://www.christies.com/about-us/welcome-to-christies

Epstein, Z. (2023). ** zive.info **. http://zive.info/

Epstein, Z., Levine, S., Rand, D. G., & Rahwan, I. (2020). Who Gets Credit for AI-Generated Art? iScience, 23(9), 101515–101515. https://doi.org/10.1016/j.isci.2020.101515

Hitsuwari, J., Ueda, Y., Yun, W., & Nomura, M. (2023). Does human–AI collaboration lead to more creative art? Aesthetic evaluation of human-made and AI-generated haiku poetry. Computers in Human Behavior, 139, 107502. https://doi.org/10.1016/j.chb.2022.107502

Nicholas, J. (2022). What exactly is ‘AI-generated art’? How does it work? Will it replace human visual artists? The Guardian (London).

Obvious. (2023). Page about – obvious. https://obvious-art.com/page-about-obvious/

Roose, K. (2022). AI-Generated Art Is Already Transforming Creative Work. The New York Times.