After-Images
AI Generated Sketchbook | 2022–2023Olive Trees, West Bank
Demolished Buildings, West Bank
Schrebergarten, Germany
Convenience Stores, Taiwan
Roadside Shrines, Taiwan
Abandoned Steel Frames of Chinese-built Monuments, Sub-Saharan Africa
After-Images is a series created using OpenAI's image generator, DALL·E 2. After some initial experiments led to intriguing results, I immersed myself in this new medium for several months during the winter of 2022-2023.
Curious to see how AI might reinterpret the ideas and subjects that have shaped both my past and ongoing work, I began applying it to my projects. Through trial and error, I developed a system of prompts, adjusting them in much the same way a photographer tweaks settings—light, focus, composition—to achieve the desired outcome.
In hindsight, the process mirrored photography itself. Just as a photographer must make numerous exposures to find a few that resonate, working with AI required generating thousands of iterations to discover a small number that were of interest.
Beyond the technical considerations, the process evoked a sense of wandering in a different kind of world—not one rooted in physical reality, but rather a vast reservoir of digital representations. Though these images are a pastiche—reconfigurations of material scraped from the internet—there was, at times, a surprising sense of presence, a ghost in the machine, as it were. Much like the revelatory collision of elements in a successful photograph, some of these AI images offered a glimpse of something new.
In all, I generated some 12,000 images. The selection presented here is a cross-section of six distinct typologies that emerged from this process.