Self-Portrait is a video-based artwork by artist Ben Cullen Williams, created with Google DeepMind Research Scientist Jason Baldridge and Google Arts & Culture Lab exhibited at Shoreditch Design Triangle Sept 16-18 for the London Design Festival, 2025. The work dives into the centuries-old question of what a self-portrait can be, using the capabilities of generative AI to explore themes of identity, technology, and expression.

Williams began by prompting Google Gemini to describe itself, which led to a dialogue that became the visual framework for the project. To define the visual world a series of models were fine-tuned using Ben’s own photographs in collaboration with Google DeepMind Research Scientist Jason Baldrige. This process allowed Google’s generative AI image model Imagen to generate images that were imbued with some of the qualities of his photographs — Williams was interested in this blurring of lines between the creator and the tool.

Hundreds of these images were then fed into Flow, where Williams sculpted them into a final, looping video work. He describes this experimental process as "akin to working in a photographic darkroom."

Find out more about this collaboration on our Google Arts & Culture Story.