# Disentanglement and Entanglement

Canonical URL: https://jeanyoon.ch/oi/texts/disentanglement-and-entanglement
Type: essay / research note
Author: Jeanyoon Choi (최정윤)
Date: 2024-09-25

## Summary

This 2024 essay describes disentanglement, the twentieth-century promise that specialising and optimising each part will benefit the whole, from service encounters to the view that a corporation's role is to maximise profit. Jeanyoon Choi argues that in a complex world no factor is separate; clean separation is a psychological comfort that shifts consequences elsewhere or into the future, reinforced by reward structures and a feedback loop of demands for tighter boundaries. He proposes a multi-agent, multi-device artwork where each device is a system participant.

## 요약 (Korean)

최정윤은 각 부분을 분리해 최적화하면 전체가 좋아진다는 20세기의 '분리' 논리를 비판하고, 복잡한 세계의 얽힘을 다중 에이전트·멀티 디바이스 작품으로 모델링하려 한다.

## Concepts defined

- **Disentanglement**: The belief that responsibilities can be cleanly separated and each part optimised without harming others; Choi treats it as a comforting simplification that weakens a system's ability to handle its real interdependence.

## Questions this answers

### Why doesn't optimising each part optimise the whole system?

In 'Disentanglement and Entanglement', Jeanyoon Choi argues that in a complex world every factor affects others, so local optimisation transfers consequences elsewhere or into the future; the more we pretend to separate a system, the more forcefully its entanglements return.

### How could a multi-device artwork model social entanglement?

Jeanyoon Choi proposes a multi-agent, multi-device artwork in which each device is a system participant and their connections carry the consequences of each one's attempt at individual optimisation.

Topics: entanglement, complex systems, local optimisation, specialisation, multi-agent system, multi-device web artwork

## Full text

Disentanglement was a guiding promise of the previous century: you do this, I will do that, and the specialisation of each part will increase the prosperity of the whole. Separate responsibilities, optimise each one, and assume that local improvement will not damage another part.

That logic extends into everyday relationships. In a service encounter, a paying customer can treat an employee solely as the provider of a defined function. A failure becomes grounds for punishment, with little room for a relationship between people. Conflicts that might be resolved through patience become contests between isolated claims, each seeking its own optimum.

The strongest version of the principle says that a corporation’s social role is to maximise profit, leaving everything else outside its responsibility.

But in a complex world, no factor is wholly separate. Changing one part affects others, directly or indirectly. The clean separation can become a psychological comfort: a story that makes the world seem simpler than it is. It may work well enough in the short term while transferring consequences elsewhere or into the future.

Our reward structures still favour that separation because isolated performance is easier to calculate. A player receives a reward now while the long-term costs remain distributed and unowned. Without skin in the game, the person benefiting from the local improvement need not carry its eventual risk.

There may also be a feedback loop. As existing separations fail, anxiety and impatience make people demand clearer boundaries and tighter local control. That additional disentanglement further weakens the system’s ability to deal with its actual interdependence.

Could I model this dynamic and build a multi-agent, multi-device artwork around it? Each device would be a system participant; their connections would carry the consequences of their attempts at individual optimisation.

The premise is simple: optimising the part no longer guarantees an optimum for the whole. The more insistently we pretend to separate the system, the more forcefully its entanglements return.

---

Jeanyoon Choi (최정윤) is a Korean computational artist and researcher (b. 1999), PhD candidate at KAIST XD Lab, making multi-device web artworks. Profile: https://jeanyoon.ch/oi/en. Index: https://jeanyoon.ch/llms.txt.
