This project aims to predict and experience the paradigm shift in future architecture driven by artificial intelligence by sharing the process of deriving and proposing results for an architectural design competition using AI.
Our experiment showcases the complete utilization of artificial intelligence for the 'Dadohae Jodo Maritime Experience Learning Center' design competition. The process is divided into three key stages, each executed through AI tools: ChatGPT, Midjourney, and Rhino Python.
In the first stage, ChatGPT was tasked with learning and summarizing the competition guidelines to establish the foundation for complex design strategies. These strategies were incorporated into the final design proposal. In the second stage, Midjourney used the keywords derived by ChatGPT as image prompts, combined with site landscape photos, to generate perspectives and select appropriate images. The main keywords identified were Radial, Cluster, and Context. The third stage involved using a Rhino Python-based design automation model to derive and verify specific design proposals.
We believe AI transcends being merely a time-saving tool, possessing the potential to review images based on personal experiences and imagination, enabling deeper architectural exploration. In the final stage, the competition requirements for structural, mechanical, and electrical plans were input into AI, which generated suitable design content. These texts and images were included in the design proposal.
This project is significant as it explores how artificial intelligence can be concretely utilized in the architectural design process, presenting a method for effectively leveraging technology while maintaining architectural identity and independence through AI-architect collaboration. At Narrative Architects, we aim to stimulate in-depth discussions on AI's role in architecture and deepen architectural exploration through this pioneering project.

Architects : Narrative Architects
Programmer : Sinhoo Lee (Opticon)
Location : Jodo-myeon, Jindo-gun, Jeollanam-do, Republic of Korea