Human-AI Collaboration in Architectural Design: A Comparative Analysis of Conceptual and Computational Form Generation

Architectural education aims to foster creative thinking and cultivate individual originality through design-based learning. In this context, the rapid integration of artificial intelligence (AI) into design disciplines necessitates an in-depth evaluation of its role in educational environments. The central aim of this study is to understand how AI interacts with conceptual design thinking and how it can influence or challenge students’ conceptual approaches to design. Developed as an experimental workshop for architecture students, the study was structured in three phases: first, students created 15x15x15 cm physical cube models; in the second phase, they developed written prompts describing their models and design concepts to generate AI-based visual outputs; and finally, they critically compared the AI-generated visuals with their original designs.The findings indicate that while AI could not replicate the depth of human creativity, it provided alternative perspectives that enriched students' design processes. The effectiveness of AI integration depended largely on the clarity and conceptual coherence of human-generated prompts. Moreover, the workshop highlighted the pedagogical value of physical model-making in design education, emphasizing embodied learning and spatial reasoning. Ultimately, this study advocates for a hybrid educational model in which AI operates not as a substitute, but as a reflective and dialogic partner in design education, enabling students to integrate intuitive and computational thinking within emerging technological contexts.