Archive Fever is an interactive visual system exploring how artificial intelligence engages with cultural memory.

The work presents a fashion silhouette that is continuously observed and classified by a machine. A style detection model assigns the image an aesthetic category while measuring its confidence in that classification.

As the system continues to analyze the image, instability begins to emerge. The visual starts to fragment and distort as a direct result of the system feeding on its own interpretation. The image begins to reference itself, entering a recursive loop where classification and visual structure collapse into one another.

The piece reflects how AI systems learn from cultural archives. Rather than generating entirely new material, they reorganize and reproduce existing references drawn from accumulated visual histories.

Archive Fever presents this process as a living system. A machine attempts to stabilize meaning while its source material gradually disintegrates.

The result is a shifting image where repetition, interpretation, and degradation become inseparable.