This project investigates the relationship between language and painting through the lens of deep learning. Drawing from art theory and personal practice, it examines how AI models interpret visual information syntactically—through form and structure—and semantically—through meaning and context.
Building on Curtis L. Carter’s view that paintings have a syntactic system akin to language, I explore how AI models process visual data. While these tools excel at reproducing structural elements, they often lack an understanding of semantic depth, particularly in emotionally or culturally rich contexts.
Artistic Methodology
I used diffusion models as a creative partner in my sketching process, treating the outputs as defamiliarizing prompts. This helped me move beyond stylistic habits and reflect on the role of machine creativity in traditional painting.
