REMNANT

Working Process – REMNANT

The project began with a broad exploration of the theme of post-digital work. I was interested in what happens when AI starts to imitate or even replace human forms of expression. My initial idea revolved around the concept of loss – the idea that voice, style, and movement could be extracted from the human and reconstructed by machines. These fragments build the AI, but what remains missing are things like memory, emotion, and individuality.

At first, the concept was wide-ranging – covering synthetic voices, AI-generated images, and deepfake aesthetics. To make it more focused, I decided to concentrate on a specific case: the synthetic voice. The core idea remained – the disappearance of human expression – but now it had a clearer and more defined anchor. In parallel, I began with early design explorations. One idea was a multi-screen setup showing the human dissolving into fragments while the AI assembled itself from those same pieces. This version was tied to the earlier, broader concept, and was eventually dropped.

With the refined focus came a new visual strategy: the human is represented with a clay-like texture – organic, soft, and unique. The AI appears in a wireframe/glitch aesthetic – technical, abstract, and hollow. These two contrasting visual elements define the entire video.

The sound layer added a second level of narration. AI-generated voices, glitches, and layered effects create tension and underline the process. The human and the synthetic seem to speak to each other – yet something essential is missing. Even if the voice can be copied, it cannot carry meaning without origin.

The final work, REMNANT, is the result of this process – a visual and sonic investigation into what remains of us when our voice becomes data, and the machine begins to speak in our place.