Phase One: Style Image

These videos showcase some of my experimental video work that uses AI as a tool within my animation and video art practice. Smell the Roses (2023) began my exploration into this workflow, focusing on how to transform a 3D fractal render into a shifting field of flowers using only a single reference image. You can find a dedicated technical explanation of that process on my dedicated project page.

Following that project, I spent time exploring other ways AI can be woven into my work. In early 2025, I leaned into using my own physical and digital artwork as the style images themselves (Elevator), and even began training a LoRA on a cross-medium collection of my art (A Trip Beyond Reality). My video Confabulations emerged from researching the psychology of memory distortion while reading “A False and Distorted Memory.” For this piece, I wanted the AI to mimic the human tendency to fill in "missing memories." I trained the model on raw video footage I had taken, mostly of my cats, and used Stable Diffusion to generate the missing pieces of those psychological spaces.

Currently, I’m working on taking the newest experimentations of training LoRA’s and using my own art as a style image within a closed system in my next project The Synthetic Memory.