전통의 '결'을, 생성하다

GENERATING THE 'GYEOL' OF TRADITION

PROPOSAL

Reimagining Korean tradition through AI: from cultural pattern to a generative system

GYEOL

OVERVIEW

Supported by the Mary Gates Research Scholarship, Gyeol explores how AI can facilitate meaningful self-reflection rather than simply generating answers. By combining qualitative research, interaction design, and human-centered AI, the project investigates how technology can help users better understand their personal values through guided reflection.

NOTION LINK —>

ROLE

Project Manager/Researcher

MENTOR

Professor Nam-ho Park

DURATION

9 Months

SOFTWARE

ComfyUI • Figma • ChatGPT API

Gyeol (결) is a Mary Gates Research Scholarship funded research project exploring how artificial intelligence can reinterpret traditional Korean visual motifs into contemporary digital textures and generative design systems. Conducted at the University of Washington under faculty mentorship by Professor Nam-ho Park, the project investigates how cultural aesthetics can be translated into computational workflows while maintaining authenticity, meaning, and historical context.

OVERVIEW

Context: Korean patterns + AI + generative systems

  • Why it matters: identity, cultural translation, design systems

  • Research question:
    → How can traditional Korean patterns be translated into generative systems using AI?

01 — IMITATION | 모방

02 — AESTHETIC EXTRACTION | 모방

03 — IMITATION | 모방

PROCESS FRAMEWORK

Understanding the structure of tradition

Understanding the structure of tradition

MAIN STEPS

01 — IMITATION | 모방


  • Pattern analysis (geometry, repetition, symmetry)

  • Manual/digital recreations

  • Key observations

02 — Aesthetic Extraction (미적 특징 활용)

Subtitle: Translating patterns into systems

Layout:

  • Process-heavy (diagrams + breakdowns)

Content blocks:

  • Motif breakdowns

  • Rule systems (grid, spacing, rhythm)

  • Parameter mapping (scale, density, variation)

  • AI workflow (ComfyUI nodes, pipeline)

03 — Expression (미적 가치 표출)

Subtitle: Generating new forms

Layout:

  • Visual-heavy (this should feel like a gallery)

Content blocks:

  • AI-generated textures

  • Variations / iterations

  • Applications (3D materials, UI, surfaces)

1 | RESEARCH