The design process began with AI-assisted research to understand the architectural and cultural significance of Gyeongbokgung Palace. ChatGPT was used to gather information about traditional Korean architecture, including roof geometry, courtyard organization, wooden construction details, and symbolic architectural elements.
This stage helped establish a strong historical foundation before starting the design development process.
After studying the palace, key architectural elements were identified and extracted as design references:
Traditional curved roof forms
Courtyard-based spatial organization
Hierarchical circulation patterns
Wooden structural expression
Korean decorative motifs
These elements became the primary design inputs used throughout the AI-assisted workflow.
AI image-generation tools were used to transform historical references into contemporary architectural concepts.
The first prompts were intentionally simple, producing generic modern buildings with weak cultural identity. Through multiple iterations, the prompts became increasingly detailed and included specific references to:
Gyeongbokgung Palace
Traditional Korean roof geometry
Cultural symbolism
Contemporary materials such as glass and steel
Sustainable architectural technologies
As the prompts improved, the quality and relevance of the generated concepts improved significantly.
Several AI-generated concepts were produced and compared. Each proposal was evaluated according to:
Cultural authenticity
Architectural quality
Functional suitability
Visual impact
Buildability
Human judgment remained essential during this stage. AI generated the options, but the final selection was based on architectural reasoning and design evaluation.
The use of AI provided several advantages throughout the project:
Faster concept development
Rapid visualization of ideas
Exploration of multiple design directions
Improved creative experimentation
Better integration of traditional and contemporary architectural elements
AI significantly reduced the time required for early-stage concept generation and allowed a broader exploration of design possibilities.
This project demonstrated that AI is most effective as a creative design assistant rather than a replacement for architects.
While AI accelerated research, visualization, and concept generation, critical architectural decisions still required human creativity, cultural understanding, and professional judgment.
The final design was therefore the result of collaboration between Artificial Intelligence and human architectural thinking.