Speaker 1

This video asks a question that sits at the heart of our digital lives: how do we appear to a machine? We are used to our own sight, shaped by memory and emotion, and to the camera's lens, with its illusion of objectivity. But there is now a third gaze, one that works through calculation. To understand it, I placed a digital version of myself—a puppet with my face and a clone of my voice—in front of various AIs.

The work documents their attempts to deconstruct a person into data: to apply a label, assign a tracking ID, map the body's motion onto a skeletal grid, and strip a voice down to plain text. My focus, however, lands on the moments of breakdown and confusion where this logic fails. For me, these are not just errors, but moments of strange beauty and surplus meaning. They are the most revealing parts of the process.

A face judged as both 'neutral' and 'fake' by two different systems at the same time is more than a technical glitch; it's a new kind of image that hints at a possible new understanding of identity. The machine's interpretation is not just a poor copy of reality; it creates a new representation with its own logic and uncanny aesthetic. These moments reflect our own unstable, contradictory digital selves.

Instead of showing only what is lost when a human is translated into data, these moments offer a new perspective. They act as a distorted mirror. In the machine's failure to see me correctly, and in the unexpected beauty of its attempt, I was given a chance to see myself in new ways.