Contemporary Dialogues Three and one photos
This project explores the photographic in the era of AI and addresses crucial questions of objectivity and representability. What aspects of a photographic message can be described through words and represented by AI? This double transformation lies at the core of contemporary dialogues between humans and machines, where words become images and images are contextualized as text by large language models.
I selected photographs from my personal projects that explore the relationship between nature and architecture, as well as contemporary urban environments and aesthetics. I prompted a GPT model with the instruction: “Provide a detailed and objective description of what is in this photo.” I then copied the resulting text and used it as input for another model to generate a new image based on the same description.
Through this process, I aimed to short-circuit the image–text transformation by closing the loop in on itself, in search of a spark that might reveal the limits of these systems’ capacities. The project is inspired by Joseph Kosuth’s One and Three Chairs, which explores a similar mechanism by presenting a chair, a photograph of a chair, and a dictionary definition of the chair in order to question the nature of perception itself.
This is an ongoing project.
