# Two-Rooms Problem: Creating inter-spatial Multi-Device Web Artwork

Canonical URL: https://jeanyoon.ch/oi/texts/two-rooms-problem-creating-inter-spatial-multi-device-web-artwork
Type: essay / research note
Author: Jeanyoon Choi (최정윤)
Date: 2025-02-07

## Summary

Jeanyoon Choi combines three directions: inter-spatial connection as a first step toward N-room systems, indirect manipulation as a source of critical distance, and circulation between semantic and nonsemantic experience. The proposal uses two rooms, a screen of overlapping generated videos, a screen of graphs and parameters, and two mobile interfaces, one for text or voice input and one with sliders and toggles. In the crossed arrangement, content entered in Room 1 generates video in Room 2, whose parameter adjustments change Room 1's display, forming a Möbius-like circuit.

## 요약 (Korean)

최정윤은 두 개의 분리된 방을 하나의 멀티디바이스 경험으로 연결하는 'Two-Rooms Problem'을 제안하며, 의미적 입력과 추상적 파라미터 조작을 교차 배치해 행위의 결과가 다른 방에서 돌아오는 뫼비우스 같은 순환 구조를 설계한다.

## Concepts defined

- **Two-Rooms Problem**: The design problem of creating one multi-device experience across two separate rooms so that an action's consequences occur elsewhere and return in another form, as the minimal case of an N-room system.

## Questions this answers

### What is the Two-Rooms Problem in Jeanyoon Choi's work?

It is a proposal for a multi-device web artwork spanning two separate rooms, the smallest step toward an N-room system that cannot be encountered from one position, where actions in one room have consequences in the other.

### Why would an interactive artwork use indirect manipulation instead of direct feedback?

Jeanyoon Choi suggests that a carefully designed indirect relationship between input and output can create critical distance, though poorly designed opacity would only confuse.

Topics: multi-device web artwork, inter-spatial installation, indirect manipulation, semantic and nonsemantic, generative video, system art

References: [Möbius strip](https://en.wikipedia.org/wiki/M%C3%B6bius_strip)

## Full text

Three directions meet in the two-rooms problem.

First, inter-spatial connection. Two separate rooms can participate in one multi-device experience. This is the smallest step towards an N-room system that no longer depends on a single exhibition space. A participant cannot encounter all of it from one position.

Second, indirect manipulation. Conventional interface design often reduces cognitive effort through an immediate, legible correspondence between input and output. I want to investigate whether a carefully designed indirect relationship can create critical distance. Poorly designed opacity would merely produce confusion or disinterest; the distance has to become part of the experience.

Third, the boundary between semantic and nonsemantic experience. We are surrounded by meaning, and AI produces and processes more of it. Yet the visible meaning of an object may matter less than the hidden parameters and relationships organising it. Rather than settle on either side, I want a circulation in which the semantic becomes nonsemantic and the apparently abstract becomes meaningful. These are working distinctions, not rigorous categories.

Typing ‘apple’ and receiving an apple video is a straightforward semantic relation. But the position of that video, the images overlapping it and its relations to neighbouring channels may eventually matter more than its subject. The shift is from the object to the system, from the part to the whole.

Proposed elements: two rooms, two kinds of screen and two kinds of mobile interface. One screen displays overlapping generated videos. The other presents graphs and changing parameters without that recognisable content. They remain synchronised. One mobile interface generates content through text or voice; the other adjusts abstract parameters through sliders, toggles and checkboxes.

Arrangement 1 pairs like with like. Room 1 contains the abstract screen and parameter controls. Room 2 contains the videos and content-generating interface. The correspondence is easy to understand, but it risks making the overall structure too self-contained for the problem I want to explore.

Arrangement 2 crosses the pairs. Room 1 contains the abstract screen and semantic input. Room 2 contains the semantic screen and abstract controls. This produces a Möbius-like circuit:

Content entered in Room 1 generates video in Room 2. That video influences how someone in Room 2 adjusts parameters. Those adjustments change the abstract display in Room 1, which in turn influences the next content input.

The crossed arrangement is conceptually stronger because an action’s consequences occur elsewhere and return in another form. It is also harder to understand. The remaining design task is to make that partial understanding engaging, through the specific content and sequence of encounters, rather than explain away the structure.

---

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.