Girl with Okappa Hair (2025) by Kay Sawatani (nom de plume) is available via lulu.com.
TEXT + IMAGES
Maximizing the Power of Story
When I received the proof copy of my first storybook, Girl with Okappa Hair, my hands were shaking with excitement. There was something special about holding the tangible product of my own creative labor in my hands. Writing has always been the medium of my choice, and the words (in both English and Japanese) included in this book are my original work. I would have never have dreamed of illustrating the book myself, though, without the image-generating AI to turn my visions into visual images.
Initially, I didn't know where this girl came from. She just kept showing up in the images I was generating with AI. It was much later when I noticed the resemblance between her and my own childhood pictures from when I was three or four years of age.
I've met quite a few skeptics who doubted that AI-image generation could be any form of creative self-expression. My experience of putting together the storybook can't be further from their perspective. The process of generating images with AI has had unexpected effects of drawing out memories and yearnings from my childhood, buried under the surface of my adult consciousness. Until I started collating and studying the collection of images for book production, I didn't even realize that:
"The quality of AI-generated images depends heavily on the human user’s ability to describe their mental image. As I search for my words and go back forth with DALL-E, the girl with okappa hair, hidden inside me, came out to play. As I looked through the generated images, I realized that every one of them had its own story" (from the Preface, Girl with Okappa Hair).
I would have never experienced this self-reflective moment, if I was randomly generating images, or trying to mimic someone else's ideas or styles. The book-making project compelled me to think about the image work I was doing with AI, and to realize the unexpected value of AI imaging.
