I've been fascinated by AI generative imagery for a while. After learning recently that I could feed one of my own images into DALL-E to use as the source, I couldn't wait to try it. Using my famous Tokyo Crow, I made 3 successive generations. Each built with the results of the previous. My work fed to DALL-E, then DALL-E's variations fed back into itself. I love how the increasingly abstract reimaginations of my work still feel related. The exploration of stripes and movement is exciting and the resulting visuals ever more haunting. At times suggestive of the original crow, but at other almost hinting at something else. To me these feel like an old duplicated VHS tape, recorded over several times creating a warping and burred reminder of what was once there. I hope you enjoy them as much as I have.