Symposium: Reimagining Generative Design

Amid ongoing fascination with the novelty and technical capabilities of contemporary artificial intelligence (AI) systems, Reimagining Generative Design challenges the notion that AI constitutes a new beginning in architecture without a past. The project connects recent architectural engagements with AI to the longer intellectual history of architecture and the technical history of the broader field of generative design, foregrounding continuity rather than a point of origin or rupture. Grounded in a review of more than 600 publications and seminal works, part of the project, Situating Machine Learning in Architecture, examines learning-based approaches by investigating their intersections with established architectural design and research lineages, showing how technological research over the past decade revisits and reframes traditions in form and typology studies, urban studies, and language-based building practices. Based on a detailed survey, the other part, A Taxonomy of Generative Design, maps salient technical paradigms and methods from the past eighty years alongside contemporary design works that mobilize them, tracing the development of generative design from knowledge-based and agent-based approaches to current learning-based approaches.