My work engages images not as stable objects, but as processes that demand legibility. I work with AI image generation, particularly diffusion models and focus on moments when images are in the process of emerging or failing to fully cohere.
AI systems require that every element conform to learned statistical structures in order to be rendered as a recognizable image. I am less interested in what AI images represent than in the conditions under which they are allowed to appear, stabilize, and circulate as meaningful. In my work as a physicist I use abstraction, mathematical consistency, and approximation as tools. That way of thinking shapes my work as an artist. Rather than treating AI as a neutral tool or as means of producing finished images, I approach it as a site where classification and normalization operate openly, and where their limits can be felt.
My previous grid paintings translated transient computational states into square grids. The AI grid is conceptually rich: it is not simply a pixel structure that displays images, but functions as the underlying substrate on which images are learned, organized, and generated. These grids embody prior visual knowledge embedded in datasets and algorithms.
In my more recent digital and painted work, the grid recedes and I focus instead on the epistemic insistence that everything must resolve into clarity in order to be intelligible. By holding onto unstable or incomplete moments, the work resists the comfort of closure and polish. I am drawn to images that are not random, images that undeniably have form, yet lack the authority required to pass as complete.
By focusing on these states, my work stays with forms that are ephemeral, dismissed and hidden by systems that prioritize efficiency and conformity.