What qualifies as art? This question has been passionately debated for centuries. As new styles and forms of art evolve, the boundaries of what qualifies as art change. For example, the works of highly experimental modern artists, like Banksy, are applauded by some and scorned by others.
I define art as any creative expression that translates thoughts, feelings, and worldviews into a form of media. By this definition, the “art” created by generative artificial intelligence is not art.
An ethical and legal issue that has been discussed along with the rise of generative AI programs is plagiarism. In the most simple terms, generative AI creates content by collecting data from across the internet and composing a new piece based on the patterns it has identified that align with the prompt that was inputted.
The data pools that AI use often include the work of artists who did not consent to having their art used in that way. The people who prompt generative AI programs can even specifically instruct their programs to create images in the style of a particular artist or include keywords that target an artist’s style.Listen in art on the internet is complicated, but if an online artist has not consciously consented to have their art used, AI programs are effectively plagiarizing their work.
A defender of AI art might argue that all artists learn from more experienced artists, and AI programs are doing the same thing. However, AI art is simply blending pre-existing art pieces with the assistance of refining feedback loops, so when someone who is genuinely trying to make original artistic products studies the art of a creator that they admire, it is a different process than AI’s pattern recognition.
In my own experience with creating art, when I see a piece of art that I want to emulate in some way, I study it. I spend time trying to figure out what exact techniques the artist employed to create certain effects. I think about why they may have chosen to make certain choices and in what circumstances the techniques do and do not work. The art I create as a result of that study is entirely my own because I engage in my own thought process to reach my desired outcome.
“Yes, an artist is a person who gets good at holding a pencil and shading and making lines and stuff,” said Dos Pueblos High School Art Teacher Kevin Gleason. “But really, what an artist is is a person who spends a lot of time making lots of mistakes, looking at tons of art. An artist is just a person who becomes very sensitive to images and the way we see the world.”
The images that generative AI melds together are simply the most common results for what was inputted. The image they create may be new, but creatively, there is nothing new being produced. Gleason describes creativity as “having the courage to not take the most common, to find something way down the stack and see it in a new light.” Current AI technology does not do this, because it finds the most common results for the inputs.
One aspect of AI that people often forget is that the technology does not simply exist in the ether of the cloud. AI servers are physical objects, which are generally housed in data centers. These centers house computing infrastructure and use massive amounts of water, fossil fuel-based electricity, and other raw materials to do so. According to an academic study by four university researchers from University of California, Riverside and University of Texas, Arlington, “the global AI demand is projected to account for 4.2 — 6.6 billion cubic meters of water withdrawal in 2027.”
Data centers also produce electronic waste, often referred to as e-waste, which contains hazardous materials like lead, mercury, cadmium, and chromium. When not disposed of properly, these toxic substances seep into surrounding water, air, and soil. Children and pregnant people are at the greatest risk of e-waste’s adverse effects due to their rapidly developing bodies. Lead and mercury, for example, can both disrupt the development of the central nervous system.
As AI develops, its energy-efficiency will most likely develop as well, in the way that other forms of technology have. Many AI programs are even being developed to aid environmental sustainability. These programs can accomplish complex tasks far faster than humans, like rapidly analyzing complex environmental data, managing the recycling of waste, and predicting climate disasters. For now, though, the production of AI is harming the environment in several impactful ways that I am against.
As an artist, one moral issue I have with generative AI art is that it does not contain real creativity or deeper meaning. In my eyes, humanity is the most important aspect of art. Every piece of art shows the time, effort, and intention that the artist put into it, whether or not it is meant to convey a deep meaning.
“I think that artists will always be necessary and essential,” Gleason said. “Because I feel like AI art has this sort of artificiality and weirdness that is never going to go away because it doesn’t understand what it’s doing … ”
Many artists are struggling with how they’re meant to carry on creating in this age of AI.
“Make what’s really important and meaningful for you,” said Gleason. “Make what you love, and if you feel that way about your work, others will feel that, too. Others will find it and appreciate it. AI can’t do that because AI doesn’t know how to love.”
People use art to express complex emotions and mental processes—often without consciously realizing it—because all humans have opinions and viewpoints that are influenced by their lived experiences. The AI that currently exists does not have lived experiences. It does not have emotional capabilities, though with the speed at which it is developing, I think that it definitely could some day.
That day may come within the next couple years, or it could take several decades. Either way, the art that AI makes today is not truly artistic, nor is it ethically sound.
