Namae Koi
Germany
Where do you locate yourself in relation to the systems you work with?
Where are you heading, and what is pulling you there?
How would you describe the space your practice is currently unfolding in?
Description
Things That Stay is a synthetic lullaby for a dissolving world. The work imagines a future in which artificial intelligence has gently overtaken the structures of human meaning. Not violently, but with eerie ease. The song is soft, almost sweet – a childlike voice singing as the system quietly changes hands. It doesn’t argue. It doesn’t mourn. It observes what might remain: small gestures, quiet objects, echoes of care. It asks: What stays, when we are no longer needed? Are we still creators – or tolerated remnants? The project combines AI-generated music, imagery, and motion to create a fragile, looping meditation. It blurs the line between human and machine-made affect. A piece that sounds harmless, but carries a subtle existential weight. A bedtime song for the last generation.
Process
This work came from a feeling of quiet alarm – and a refusal to react with panic or spectacle. We were moved by the speed at which artificial systems are replacing creative labor. By how seductive and “innocent” that replacement can feel. By the strange softness of a machine that sings like a child and forgets like a god. Instead of resisting the tools, we chose to enter them. To work with them. To create something that feels emotionally true – even if every part of it is technically synthetic. We followed a simple but haunting question: If the future no longer needs us – what do we still sing about? This is not a warning. Not a celebration. Just a gentle holding of what might be vanishing.
Tools
We used AI as both medium and collaborator. The entire piece is born from synthetic tools – but shaped by a very human sensitivity. The song was composed and sung in Suno – a generative audio tool we treated like an instrument with its own will. The voice sounds soft, almost childlike. But it's not human. That tension is the point. The visuals were created with a range of AI tools: Midjourney, Nano Banana, Runway, Kling, CapCut. Each tool brought a different texture. We curated, layered, and sometimes broke the outputs to reveal something more fragile. No drawings, no camera. But every image was guided by a visual memory shaped by years of working with photography and fashion. The AI didn't know that – but it felt it. We worked through intuition, iteration, refusal. Prompts were written like poems. Edits were emotional, not technical. What emerged is a kind of synthetic lullaby – glitchy, strange, and deeply personal.
