Cultivating reflective designers: process-oriented pedagogies for harnessing AI’s creative unpredictability in architectural studios
This study examines how AI image generation tools influence design thinking, creativity, and self-reflection in architectural education. Using a 15-week integrated studio with fourth-year undergraduates, it analyzes AI engagement and cognitive shifts across conceptual, developmental, and detailed design stages through mixed methods, including Creativity Support Index (CSI) analysis and comparisons between student self-assessments and tutor evaluations. The studio adopted a process-oriented pedagogy emphasizing iterative feedback, stage-based design logs, and a collaborative whiteboard platform. Leveraging AI’s capacity for low-cost failure – rapid, low-risk experimentation with multiple alternatives – it accelerated experimental loops and reinforced the cycle of outcome – interpretation – problem redefinition. AI fostered divergent idea generation, nonlinear thinking, and externalization of concepts while challenging conventional typologies of program, form, and materiality. A key pedagogical aim was to channel AI’s unpredictability into productive learning opportunities. Over time, students shifted from perceiving AI as uncontrollable to engaging with it as a design partner requiring interpretation and adaptation. Findings suggest that AI integration under this framework enhanced self-reflection, critical thinking, and creative decision-making. The study repositions AI from a mere visualization tool to a cognitive catalyst that broadens design thinking and supports co-discovery, proposing strategies for cultivating reflective designers in AI-integrated studios.
Authors
- Eunki Kang (ORCID: https://orcid.org/0000-0002-8920-3976)
- Eun Joo Park (ORCID: https://orcid.org/0000-0002-9176-991X)
- Yoon-jeong Shin
Institutions
- Soongsil University (KR)
- Sejong University (KR)
- Sungkyunkwan University (KR)
Publication Details
- Journal
- Journal of Asian Architecture and Building Engineering
- Published
- 2026-01-29
- DOI
- https://doi.org/10.1080/13467581.2026.2621521
- Primary Topic
- Architecture and Computational Design
- Type
- article
- Field-Weighted Citation Impact
- 0.00