Façade Psychology Is Hardwired: AI Selects Windows Supporting Health

This study uses generative AI to investigate the influence of building façade geometry on human physiological and psychological health. Employing Christopher Alexander’s fifteen fundamental properties of living geometry and a set of ten emotional descriptors {beauty, calmness, coherence, comfort, empathy, intimacy, reassurance, relaxation, visual pleasure, well-being} in separate tests, ChatGPT 4.5 evaluates simple, contrasting window designs. AI analyses strongly and consistently prefer traditional window geometries, characterized by symmetrical arrangements and coherent visual structure, over fragmented or minimalist–modernist alternatives. These results suggest human cognitive–emotional responses to architectural forms are hardwired through evolution, privileging specific geometric patterns. Finally, ChatGPT o3 formulates ten detailed geometric rules for empathetic window design and composition. It then applies these criteria to select contemporary window typologies that generate the highest anxiety. The seven most anxiety-inducing designs are the most favored today worldwide. The findings challenge contemporary architectural preferences and standard window archetypes by emphasizing the significance of empathetic and health-promoting façade designs. Given the general suspicion among many readers of the frequently manipulative and unreliable use of AI, its use in this experiment is not to validate design decisions directly, which would put into question what the AI is trained with, but to prove a correlation between two established methodologies for evaluating a design. AI is used as an analytical tool to show that Alexander’s geometric rules (the guidelines proposed beforehand) closely match emotional reactions (the desirable outcomes observed afterward). This novel use of AI suggests integrating neurodesign principles into architectural education and practice to prioritize urban vitality through psychological well-being.