Generative AI and Extended Reality in Collaborative Architectural Design Education: An Exploratory Studio Study
| Source: arXiv AI
Tags: generative AI, extended reality, architectural design, education, human-computer interaction, design tools
A 27-student classroom experiment found that combining generative AI with collaborative XR for architectural design reduced student confidence and outcome expectancy versus a control group, while panel-rated design outcomes were not significantly different — suggesting GenAI+XR may change the design process without improving end products.
Details
GenARch integrates generative AI image generation with a shared XR (extended reality) environment for collaborative architectural design, deployed in an undergraduate design studio. Twenty-seven students in seven teams participated; four teams used GenARch, three continued without it. Quantitative results show larger declines in design self-efficacy and outcome expectancy for the GenARch condition. Perceived conflict management improved in the GenARch group (positive difference-in-differences estimate). Panel-rated final design presentations were not significantly different between conditions. The study is exploratory and mixed-methods, with a small sample that limits statistical power. Qualitative interviews offered complementary insights: GenAI supported idea externalization and rapid visual reference generation, while XR provided spatial and scale-based evaluation. Students flagged challenges including control limitations (AI outputs diverging from design intent), dimensional fidelity issues, and shared attention difficulties in the XR environment. This is an exploratory pilot study (N=27), so findings should be treated as hypotheses for future work rather than conclusions. The result that confidence declined while outcomes stayed comparable mirrors patterns seen in early adoption of other creative tools.