‘At the ice of knowledge will he yet freeze to death!’
— Friedrich Nietzsche, Thus Spoke Zarathustra

Cooling Down is an in-progress lecture performance combining video projection, AI-generated ‘slop’ visuals, live desktop navigation and physical storytelling.

The world is heating up, and we are building systems to cool it down. We cool servers, databases, images and bodies in order to preserve the things that contemporary systems are built to extract and retain: memory, identity, behaviour, prediction and value.

In data centres around the world, cold air keeps information stable and machines running. In cryogenic fantasies, this same logic extends to the body, as if life itself could be paused, stored, and resumed at a later date. Cold becomes more than just a temperature. It becomes a method of survival.

The lecture performance follows two contemporary figures of technological immortality: Larry Ellison and Alexandr Wang. Ellison imagines survival through the extended body: anti-ageing, cryonics, repair. Wang represents another form of survival: prediction, annotation, alignment, and scale. One tries to delay the end of the body. The other tries to reduce the uncertainty of the future.

In contrast to these fantasies of continuation, the project returns to telos: purpose, direction and ethical end. As in the story of Patroclus, the dead body is not a failed machine, but rather one that has a passage to complete. However, cryonics and data immortality interrupt this process, turning burial into storage and mourning into maintenance. The dead become material waiting for reactivation.

Set against the backdrop of Scandinavian cold infrastructures, including hydropower, server farms, fan noise, cooled air and emotional labour, Cooling Down asks who benefits from ‘forever’, who pays for it and what is lost when meaning is replaced by maintenance.

Cold appears as care, regulation, distance and refusal. It cools the machine, but also responsibility.

❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍

AI Song Energy Count

This piece serves as a quantitative log of the computational resources required for its synthesis. It provides a sterile, data-driven account of the 0.045 kilowatt-hours of energy consumed and the 162,000 joules of thermal output generated by high-density GPU processing.

The composition focuses on the metabolic cost of AI music. It ends precisely when the energy-to-audio conversion is complete.

Made in Suno-AI

Lyrics:

Play ↓

Youtube link

〄