Recipe for AI integration in architecture : learning from good practice precedents

Abstract

Artificial Intelligence (AI) is transforming all disciplines, impacting research, practice, academia, and industry. Architecture is no exception, yet architects have not fully embraced the digital paradigm. AI tools and methods present an opportunity to enhance the positioning of architects within the evolving digital landscape and the increasing complexity and sustainability demands of architecture. The study proposes a rules-based conceptual framework for an AI-driven, procedureoriented approach to the design process, informed by historical and contemporary examples of embedded design knowledge as good practice. The recipe for AI integration is derived from an educational methodology illustrated through student projects that combine analogue techniques with digital principles to explore the interplay between algorithmic thinking and AI with a focus on architectural geometry. The paper demonstrates how intuitive investigation, iterative learning, and collaborative processes support the development of computational thinking by embedding design knowledge into procedural systems – introductory for future application. Finally, the paper discusses key guidelines for how AI can be meaningfully integrated into the discipline of architecture.