Beyond the Paper Screen / Artificial Intelligence Can Open Up Novel Opportunities for Creative Expression

In some ways, the Japanese word kiyo is like “handy” or “crafty” in colloquial American English, connoting the ability to use one’s hands and make or fix things. But in kiyo, there’s a bit more emphasis on dexterity, the ability to do intricate work using not our hands per se, but just the tips of our fingers.

My maternal grandparents were both traditional artisans, who left home at a young age and apprenticed under a strict master in their respective trades: cabinetmaking for my grandfather and kimono-tailoring for my grandmother. To them, being kiyo was their livelihood. They praised me for everything I did, as any grandparents would, but they never called me kiyo. A girl who couldn’t sew straight or paint neatly within the lines was just the opposite of it: “bukiyo” (literally “lacking kiyo”). So, I grew up thinking that any artsy-crafty things that required dexterity were not for me.

It was an eye-opening experience to take “Introduction to Ceramics” in my very last semester in college. Our professor, whom everyone called by his first name, Nils, taught this class not to make us kiyo at pottery-making, but to show us how to explore and “play” with clay. I took his lead and plunged in. I learned to make well-shaped cups, plates and vases, but they never stayed that way. I pushed and stretched them, scraped them with a plastic knife from the cafeteria (one of my favorite tools, it turned out), punched a hole in them when they were leather-hard. Nils didn’t say much but quietly encouraged my experimentation. One day I went into the studio and found an open artbook in my work area, with a photograph of Japanese pottery that I was unsuccessfully trying to emulate. So, I was tickled when he teased me one day, “Can’t you make anything straight for once, Sawa-chan?” That was the first time in my life when someone validated my bukiyo way.

It was inevitable that ceramics went by the wayside through the years of getting through graduate school and building my academic career as an anthropologist. But I never forgot the simple pleasure of creating something — however small, simple or imperfect — of my own. Anthropology also taught me that making things is a quintessential human trait. To make things and to share what we have made is to experience our own humanity.

Recently, the tale of this (formerly) bukiyo girl took an unexpected turn: the image generating AI called DALL-E. I heard about DALL-E shortly after I began my journey to explore the academic use of ChatGPT and other text-generating AIs. At first, I was disappointed by the images I was generating, which looked nothing like what I saw other (more experienced) users posting online. But then I discovered an online community where I met dozens of accomplished artists experimenting with this new form of creativity. I was amazed at how generous and helpful they were to complete newcomers like me, sharing their knowledge freely and answering rudimentary questions. So, like when I started Nils’ pottery class, I plunged right in. Soon I began to learn how to write more effective prompts and generate high-quality images that are closer to my vision.

There is a great deal of ambivalence about the role of artificial intelligence in society. The questions of originality make many wary of even giving generative AIs a try, for fear of inadvertently violating the intellectual property rights of others. In fact, earlier models of DALL-E and similar image-generating AIs caused a stir by allowing users to generate fake images of living public figures or to make exact replicas of works by living artists.

By the time I got into image generation, DALL-E was already smarter, and continues to become better about maintaining strict ethical standards. Some still wonder whether an image that a human user generates with AI can be called “original” or qualify as “art.” These are complex questions that no one person can answer definitively; rather, society as a whole will have to come to terms with the blurred human-machine boundaries and discern the value of the products that come out of our collaboration with AI.

In the meantime, I am trying to imagine what my late pottery teacher would have said about this latest turn in artistic expression. His entire life centered on creativity, to turn his ideas into something concrete, that which can be seen, touched and shared with others. And yet, he surprised me once when he said: “There’s nothing new under the sun, Sawa-chan. We think we are so creative, then we find out someone else already thought of it.” He admired the ingenuity of those who came before us and knew how rare it was for us to come up with anything entirely original. But it is in our effort to push the boundaries that we have a hope of adding just a little something to that human history.

A few weeks into using DALL-E, others in my online community began to take notice of my unique style, which reflects a unique combination of cultural and personal influences that have shaped my life. I can also see individual styles in others, too. How we each choose from a vast range of options made possible by the AI’s storehouse of visual data can only be called distinctive and personal. I am already beginning to see others trying to emulate my style. In doing so, they are also on a path to developing their own.

Sawa Kurotani

Kurotani is a professor of anthropology at the University of Redlands.