# How to build an evolving interactive system art?

Canonical URL: https://jeanyoon.ch/oi/texts/how-to-build-an-evolving-interactive-system-art
Type: essay / research note
Author: Jeanyoon Choi (최정윤)
Date: 2025-03-18

## Summary

In this 2025 essay, Jeanyoon Choi asks how interaction can change an artistic system's nature rather than simply navigate its preset possibilities. He compares this to designing an active scientific experiment, using a hypothetical study of preference for the golden ratio: choosing among images, then adjusting a slider, then a third, freer level that someone would choose to experience for its own sake. He considers hot–cold games, binary search, adaptive psychophysical procedures and other inputs, and imagines two outputs: a research result and an artistic experience, as neural-network visualisation relates to SoTA.

## 요약 (Korean)

최정윤은 이 글에서 관객의 상호작용이 시스템의 규칙 자체를 바꾸는 진화형 인터랙티브 시스템 아트를 묻고, 황금비 선호 실험 설계를 비유로 삼아 연구 결과와 예술적 경험이라는 두 가지 산출물을 상상한다.

## Questions this answers

### How can an interactive artwork evolve through audience participation?

In 'How to build an evolving interactive system art?', Jeanyoon Choi argues that the key question is whether an action changes the system's possibilities and rules, not just its parameters within a fixed axis, as a slider does.

### How does Jeanyoon Choi relate scientific experiments to interactive art?

He suggests that designing an active experiment, such as a study of golden-ratio preference, parallels designing active interactive art: both need situations where someone genuinely explores rather than performs a prescribed gesture. Such a project could yield both a research result and an artistic experience.

Topics: system art, interactive art, experiment design, golden ratio, neuroaesthetics, adaptive psychophysics, interaction modality

References: [Golden ratio](https://en.wikipedia.org/wiki/Golden_ratio), [Psychophysics](https://en.wikipedia.org/wiki/Psychophysics), [Binary search](https://en.wikipedia.org/wiki/Binary_search_algorithm), [Generative adversarial network](https://en.wikipedia.org/wiki/Generative_adversarial_network), [Brain–computer interface](https://en.wikipedia.org/wiki/Brain%E2%80%93computer_interface), SoTA, [Apple Vision Pro](https://en.wikipedia.org/wiki/Apple_Vision_Pro)

## Full text

How can interaction change the nature of an artistic system, or even generate a new one, rather than simply navigate its existing possibilities? This is the question behind an evolving interactive artwork.

Designing an active scientific experiment offers a useful parallel. In both cases, I need to create a situation in which someone genuinely explores, rather than obediently performs a prescribed gesture.

I distinguish explicit semantic controls such as buttons and GUI options; implicit bodily inputs such as motion, touch, expression or a brain–computer interface; language through text or voice; and ambient environmental data. A nonsemantic action can be transformed into a meaningful system input. But the important question is not only the modality. It is whether the action changes the system’s possibilities.

Consider a hypothetical study of preference for the golden ratio. At one level, a participant chooses among two to six images. At another, they adjust a slider until the ratio feels right. The slider is more active, but the participant still moves along a fixed axis while the rules remain untouched.

What would a third level look like? I want an experience someone would choose for its own sake, perhaps even pay to encounter, rather than complete because an experimenter instructed them to. They should explore with pleasure and have enough freedom for unexpected behaviour to matter.

The change from manipulating ten GAN sliders to generating an image through language suggests how a new interface can expand the space of possible actions. Yet the experimental question here would still concern proportion, so that freedom needs a relation to the ratio.

Possible approaches include a hot–cold or higher–lower game; binary search; adaptive psychophysical procedures; spoken feedback; or two simple buttons that gradually narrow a range. I am also interested in the way a movement overshoots a small target, reverses and gradually converges. Could that become an experience rather than merely a measurement technique?

Other inputs could include phone motion, hand tracking, a headset or gaze. Gaze raises a particular design problem if it is both the control and the variable being measured. The study would need to distinguish the two roles. The stimulus itself might move between abstract and descriptive forms. A Fibonacci-like process of narrowing is another possibility.

I imagine two outputs from such a project. One is a research result. The other is an artistic experience that grows from the experimental situation without remaining an explanation or demonstration of it.

Neural-network visualisation relates to SoTA in this way. What would occupy the equivalent position beyond a golden-ratio experiment? The answer requires more than a narrative laid over the study. It requires a system and a form of interaction through which the question becomes an experience.

---

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.