The Soul of the Machine
How Artificial Intelligence's Imitation of Art Exposes the Exceptionality of Mankind
The rise of artificial intelligence has raised new questions of identity. What does it mean to be human? If machines can perform what we considered to be a solely human action, what does that say about who we think we are? Over the past two years, this question has (and still does) fascinate me.
For my Senior Thesis (i.e. my capstone paper and defense), I decided to dive deeper into this question, specifically in its relationship to art. The rise of artificial intelligence forces us to reexamine why we create. Much of the existential dread about AI surpassing us lies in the subtle assumption that we ourselves are simply biological machines.
This assumption has been a core doctrine to the evolutionary naturalist who asserts that all human behavior can be reduced to biological processes adopted for survival. Artificial intelligence, which was inspired by this reductionistic model of man, allows us to test the validity of this assumption. If our actions are simply the outputs of biological algorithms, then a machine should be able to perform the same functions.
By comparing the way machines create to the way we create, I believe we get a closer glimpse at what sets us apart from other critters and creations. For this piece, I’ll focus on three main areas: how we understand versus how machines “understand,” how our desires compare to problem-solving algorithms, and how these two characteristics combine in our creative drive.
How do Machines Work?
The question of whether chatbots or other models truly understand what they are saying is a fascinating one. At first glance, it seems obvious to us that something which can communicate ideas about the history of art, explain how to take the volume of a rotated function, or draw you a funny picture of a cat must know something about what it produces. However, here begins a theme we will hit on throughout this piece: just because something works doesn’t mean it is performing anything beyond imitation.
Large language models (LLMs) like Chat-GPT work by gathering a large selection of text and finding statistical patterns across it. They use these patterns to generate the next word in their replies. They are no better than the little boy bluffing his way through a book report on a book he never read. He may look at the title of the book or its description and attempt to guess what the book was about from that, but, if he is correct at all, it’s not in part to him actually understanding the book. As he listens to other students speaking about the book, his guesses will get closer to the true content of the book. While his bluff becomes more accurate, it is still a bluff. The same goes for Chat-GPT. The functionality of an LLM is directly related to the quality and quantity of its data.
However, human comprehension is conceptual in nature. When I am speaking with a friend, I use words to represent certain ideas. If I say the word “ball” or “dog,” you can probably close your eyes and imagine a ball or dog. Your mind has taken its sensory experiences of a “ball” and “dog” and formed it into a concept that you can contemplate and imagine.1
Why and How We Create Art
It’s out of this understanding that we are able to create and reflect on art. Shifting gears a little bit, let’s think about why we desire to see and create art. Some theorize that the ability, time, and resources it takes to create art are signs of wealth and power. To be an artist (or, really, to have the ability to try to be an artist) is an evolutionarily desirable trait. Others believe art arises from the cultural myths that connect societies together. In other words, there isn’t really a pragmatic benefit for art besides playing into said myths. Many of these theories explain the desire for art as if it were a pragmatic impulse aimed at survival.
However, I believe these origin stories for our desire for art, especially our desire for beautiful art, narrow the scope too much. If these arguments are correct, they would perhaps explain our desire to build large and useful structures like food plants, factories, and apartments. But, why must these structures be beautiful? Why must it be beauty that fires our dopamine receptors and stirs our hearts? After all, beauty is often at odds with utility and efficiency. Yet, many still assert, like Victor Hugo’s Bishop, that “[t]he beautiful is just as useful as the useful. Perhaps even more so.”2
Of course, some would probably reply to me that this “irrational” impulse is a clear example of the power of cultural myths. As author Yuval Harari writes of the building of monuments like Göbekli Tepe, “Why would a foraging society build such structures? They had no obvious utilitarian purpose.”3 In a world filled with disease, destruction, and death, there seems no pragmatic purpose to set aside the task of survival to create beautiful monuments. However, the creation of beautiful art is a phenomenon we see across cultures. If we assert that beautiful art is simply the result of cultural myths, we are left at a perplexing point. Either the entire world has cross-culturally been captivated by a millennia-old lie that beauty is worth the sacrifice, or, perhaps, our theories for why we have the desire for beauty are too small. Perhaps, there is a reason behind mere utilitarianism. We create art from our conceptual understanding of the world and our desire to express ourselves; both of which cannot be adequately summarized by simple pragmatism.
Conclusion
The worldview that defines us as chemicals divorces us from meaning. To the materialist, the longing for meaning is only a wishful dream we cling to to cope with the chaos that often surrounds us. Artificial intelligence’s advance into art seemed to confirm this. If machines could imitate our soulful creations, perhaps our longings were just as artificial. But, as we have explored, machines can and will never understand as we understand, feel as we feel, or create as we create.
Once we awaken from the dogmatic slumber of materialism, we are allowed to wonder if the meaning we ascribed to our tears was more than myth. Materialism can’t account for our longing to use the scars and shards of our lives as instruments to restore broken places with beauty. Art is often shaped by tears. As painter Edvard Monk reflected, “What is art? Art grows out of grief and joy, but mainly grief. It is born of peoples' lives.”4
Despite the death and darkness that often fill our lives, we are not playing into silence. Only now, since we have been forced to remember why we created the music, do we recall the reason the piano keys resonated with our hearts. The notes echo a long-forgotten hymn of hope. The world in constant change maintains a melody, and we are playing in harmony.5
References
For a very for interesting and enlightening dive into this, see Percy Walker’s chapter on semiotics within Lost in the Cosmos: The Last Self-Help Book and Owen Barfield’s Saving the Appearances.
Victor Hugo, Les Miserables, trans. Christine Donougher, Deluxe (New York, New York: Penguin Classics, 2015), 25.
Yuval Noah Harari, Sapiens: A Brief History of Humankind (United States of America: Harper Perennial, 2015), 90.
Ragna Stang, Edvard Munch: The Man and the Artist (London: The Gordon Fraser Gallery Ltd., 1979), 15.
A paraphrase of Boethius’ poetic opening, “a world in constant change maintains a harmony,” from Book II of Consolation of Philosophy.
The Consolation of Philosophy, trans. Victor Watts (New York, New York: Penguin Books, 1999), 45.