Students used GenAI and XR in complementary roles: GenAI for ideation and visual reference generation, XR for spatial, contextual, and scale-based evaluation¶

Claim · All claims
Evidence · 1 study · causal r1 · q1
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q1 i? GenAI primarily supported ideation and visual reference generation, while XR enabled contextual and spatial understanding, including perceiving architectural elements at realistic scale. → Xiao 2026

Evidence¶

Xiao 2026¶

Xiao, Y., Chen, M., Li, Y., Powers, N., Wiesenfeld, M., Smith, G., Farzin, S., Liu, S. (2026). Generative AI and Extended Reality in Collaborative Architectural Design Education: An Exploratory Studio Study. arXiv preprint arXiv:2609.13494. https://arxiv.org/abs/2609.13494 doi:10.48550/arxiv.2609.13494

q1 · i? · causal · r1

Thematic analysis of semi-structured group interviews with the four GenARch teams (~35 minutes each), following Braun and Clarke's reflexive six-phase approach. Students reported that full-scale embodied viewing gave "a clearer sense of proportion and spatial presence compared with screen-based representations," while GenAI images externalized rough conceptual ideas.

"GenAI primarily supported ideation and visual reference generation, while XR enabled contextual and spatial understanding. In particular, XR helped students visualize the site landscape and perceive architectural elements at realistic scale."

Discussion¶

Related Claims¶

• The perceived value of GenAI and XR varied across design stages: GenAI suited early exploratory prototyping, XR once a 3D artifact existed for spatial evaluation — possibly the same claim (merge candidate)
• Interviews indicated flipped learning supported studio readiness while the GenAI module served as a supplementary tool for early-stage visual ideation — related
• A bespoke GenAI image workflow most robustly expanded architecture students' creative search space, while less uniformly supporting refinement of final design decisions — related
• Students reported challenges with generated-model scale fidelity, shared attention in multi-user XR, motion comfort, and control over generation — related
• Teachers' tool preferences are stage-dependent, with pre-gen valued for ideation, mid-gen for structuring, post-gen for finalization, motivating hybrid workflows — related
• Students rated the flipped-learning component highest and the GenAI module lower, with GenAI rated stronger for ideation than for time saving or direct output use — related
• Prompts referencing architectural styles or local architectures yielded less desired image outcomes than future-themed prompts, motivating selectable checkpoints in GAI-A — related