Art, Artificial Intelligence, and Pooping Rainbows
By Kevin Si
If you had told me twenty years ago that the year 2023 would be encapsulated by artificial intelligence, massive global conflict, and tumultuous domestic politics, I wouldn’t have had any reaction. Probably because I wasn’t alive twenty years ago.
Nevertheless, the surge of language models like Chat-GPT and text to image models like Dall-E have necessitated a similarly rapid discussion of the ethics and impacts of these models. Most notably, the development of models that are able to produce compelling and highly detailed pieces of art from a couple keywords have garnered ire and admiration from many in the artistic and technological community. Many believe that AI art poses serious questions regarding theft of intellectual property, job competition, and artistic integrity. Emad Mostaque, the founder of Stability AI (a company behind a text-to-image model) defended AI-generated art that would “poop rainbows” to alleviate a “constipated” art industry.
However, before the argument takes shape, it is important to define exactly what is being argued. Art is one of the most heterogeneous fields, with a nearly infinite number of mediums that could be considered art. Music, architecture, paintings, digital art, and pottery are all widely considered mediums of art and yet have vastly different backgrounds and methods of creation. Artists in the traditional oil-on-canvas sense like Picasso and Van Gough simply defined art as something that intentionally expressed the author’s emotions, experiences, and messages. If their definition of the field is accepted, then, by all means, what AI creates nowadays is most certainly not art.
Literally, the techniques and mediums that AI uses to generate art is the same as human generated, but it lacks the meaning and message that human-generated art wields. An artist’s emotional tie with their artwork is what makes it a piece of art. Artificially-generated relies on algorithms, mathematics, and existing images to create its images. The surface-level tone and emotion in many AI-generated artworks are only cheap mimics of unique and original human experiences and emotions in art the AI cobbled together.
While the classification of AI generation as art or not is relatively straight-forward, the usage of AI in art raises many multi-faced arguments. Most commonly, AI products put forward as one’s own work in competitions, the workplace, or for sale leads to much contention. For example, an AI-generated art piece won first place in a Colorado art contest, and the SAG-AFTRA strike was partially motivated by the usage of AI artwork, voice generation, and writing in Hollywood. All creatives, including traditional artists, are threatened by the rise of AI. Unfortunately for creatives, there are valid and persuasive arguments on both sides of the debate. Many developers and supporters of AI development cite important facets of technological advancement, job expansion, and efficiency and automation. On the other hand, opponents argue against AI, citing job competition, a lack of human authenticity, and bias. In the artistic field, the biggest moral qualm against AI-generated artwork is the lack of human authenticity.
The path forward as a society dealing with the impact of AI in the creative fields is straightforward, for now. Many artists have adapted to use these softwares as creative aids to help establish baselines for their work so that they can focus on the more intensive and demanding creative aspects. Usage of AI, for now, should remain as a tool and not the sum total. Regardless of how much work an “artist” may put into generating their artwork, it is unethical to have no personal creative input into a work of art and compare it to human generated artwork. Ergo, in this sense, Mostaque was right. Generative AI’s rainbow defecation might just be what artists and creatives should embrace in this modern era of technology.
