Stable diffusion in architectural design: Closing doors or opening new horizons?

This study explored integrating Stable Diffusion, a generative artificial intelligence (AI), into architectural design workflows, focusing on its impact on design process and students’ learning experiences. A comparative analysis revealed an optimized workflow incorporating Stable Diffusion, which e...

Full description

Bibliographic Details
Published in:International Journal of Architectural Computing
Main Author: Cao Y.; Abdul Aziz A.; Mohd Arshard W.N.R.
Format: Article
Language:English
Published: SAGE Publications Inc. 2024
Online Access:https://www.scopus.com/inward/record.uri?eid=2-s2.0-85200681509&doi=10.1177%2f14780771241270257&partnerID=40&md5=73af20e7d71bc8bb886aea0911004bc7
Description
Summary:This study explored integrating Stable Diffusion, a generative artificial intelligence (AI), into architectural design workflows, focusing on its impact on design process and students’ learning experiences. A comparative analysis revealed an optimized workflow incorporating Stable Diffusion, which enhanced design exploration, conceptualization and visualization in early design stages. Demonstrations showcased text/image-to-image capabilities generating architectural visuals. Employing a mixed-methods research design, which encompasses comparative analysis and a thorough questionnaire-based exploration, the research sheds light on the challenges and opportunities of integrating Stable Diffusion into architectural education and practice. While receptive, some concerns existed around AI automation risks. The paper contributes to a deeper understanding of the transformative potential of generative AI, particularly Stable Diffusion, in reshaping workflows and educational dimensions of architectural design. Findings advise integrating emerging AI like Stable Diffusion into architecture curricula to equip students for AI-driven industries, emphasizing judicious human-AI collaboration. Further research could continue optimizing hybrid human-AI design workflows. © The Author(s) 2024.
ISSN:14780771
DOI:10.1177/14780771241270257