in transitu live, 2022-2026

in transitu live is a live performance with/against gender classification "artificial intelligence".

The performance lasts between 30 and 60 minutes and consists of me sitting in front of my laptop having a bodily and verbal conversation with a gender classification algorithm.

The feed from the computer's webcam is projected onto a wall behind me, so spectators can follow along with the live gender classification results.

During the performance I use different techniques to transform my algorithmically perceived gender into different categories. Some of the techniques are based on makeup, others are based on body language and some are reliant on facial expressions.

Throughout the live performance, I talk to the algorithm out loud about what gender is and how it relates to gender expression, data, machine learning, computers and human experiences.

in transitu live is a direct result of my work with the Instagram project in transitu. It comes out of a desire to investigate more directly how gender classification algorithms can be tricked into seeing other genders.

Whereas the Instagram project is slow, confounding and investigative, the live performance is fast, controlled and revealing. Through experimenting with this specific gender classification "artificial intelligence", I have learned exactly what techniques work to convince the algorithm to see something different than it's initial guess. I show these techniques to my audience, thereby exposing all algorithms and "artificial intelligences" as something trickable.

in transitu live was performed for the first time at Debutfest in 2022. The gender classification algorithm face-api.js is the basis of the custom software that I've written for the performance. Images from the Debutfest performance used on this page are taken by Niclas Hawkesworth. Images from the KP25 performance used on this page are taken by Semih Simsek.