Exploring the Impact of AI Chatbots on Architectural Design: A Case Study in an Architecture College Design Studio
Abstract
The advent of chatbots has provided designers and students with a platform to generate architectural design images swiftly and effortlessly. However, the extent to which these images aid or confuse users, remains a pertinent question. This research presents a case study conducted in a design studio course at an architectural college in Hong Kong, aiming at the design of a single-family house. Six users were instructed to interact and utilize the chatbot “Stable Diffusion” to generate conceptual images during the schematic design phase. Observations revealed four distinct approaches employed by users: (A) generating fully designed buildings with the chatbot, (B) producing inspiring images reflecting users’ sensations, (C) developing abstract images to inspire the housing project, and (D) a control group where students did not use the chatbot and followed the traditional design approach. In Approach A, users rapidly produced realistic, fully designed houses with different design styles to gauge clients’ preferences, albeit at the expense of deeper design exploration. Approach B saw students generating diverse imaginative images to convey their design sensations, offering a visual aid in conceptualization. While these images exhibited minimal resemblance to the final design, they served as valuable visualization tools. In Approach C, users generated abstract art images, which continued to influence architectural concepts throughout the entire design project, resulting in designs bearing moderate to major resemblance to the initial chatbot images. In conclusion, users adopted various approaches in utilizing chatbots, with Approach C proving beneficial throughout the design process. These findings demonstrate the potential for AI chatbots to assist designers in creating conceptual images during the schematic design stage.
Author: Kevin Yim
Published in: Canada International Conference on Education, 2024
- Date of Conference: 23-25 July, 2024
- DOI: 10.20533/CICE.2024.0040
- Electronic ISBN: 978-1-913572-65-5
- Conference Location: Toronto Metropolitan University, Toronto, Canada