Pedagogy 2025

Abstract

This article presents a pedagogical framework for integrating Artificial Intelligence (AI) into performance-driven architectural design education. The framework combines a graduate research design studio with a research seminar, creating an experimental platform that bridges data-driven analysis, generative exploration, and theoretical inquiry. Through this approach, students critically engage with AI as both a design medium and a research methodology. AI is embedded in design and physics-based simulation workflows using a range of computational techniques, including machine learning, computer vision, reinforcement learning, and generative modeling. Students explore form-finding, spatial organization, and façade configuration while optimizing energy efficiency and environmental performance as interconnected design processes, operating across scales from architectural components to buildings and neighborhood systems. By coupling hands-on studio experimentation with seminar-based critical discourse, the course cultivates dual literacy: technical fluency in AI-driven performance modeling and reflective understanding of the epistemic, ethical, and aesthetic dimensions. This pedagogical approach positions AI not merely as automation but as augmentation, enabling new forms of reasoning, multi-scalar thinking, locally sensitive, and data-informed creativity. The article outlines the course structure, learning outcomes, lessons learned, limitations, and broader implications for architectural education, offering a replicable model for embedding AI within performance-based design pedagogy. It shows how integrating AI into architectural practice and education can foster a design exploration that is adaptive, analytically rigorous, and critically attuned to emerging technological and societal challenges.