Book
Starting with one photo, I used DALL-E 2 to create four variations. Following a predefined algorithm, I selected one variation as the next input and repeated this nine times, producing 36 variations per algorithm. The process was repeated across four algorithms, resulting in 144 variations and nine "base" photos per segment.
To add context, I used Jina AI to generate descriptions for the nine base photos. These outputs, along with the visual data, were analyzed through three lenses: location, content, and affect. This analysis was inspired by Rorschach’s inkblot tests, which interpreted psychological responses to abstract visuals.