From Frankenstein to Gulliver urbanism: Reframing AI and scale in urban theory

Artificial intelligence (AI) is increasingly embedded in urban governance, infrastructure and design, yet urban theory still lacks a clear language for analysing the scale asymmetries through which these systems operate. This article develops Gulliver Urbanism as a conceptual framework for understanding how AI reshapes urban power when it processes information, acts, and coordinates across cognitive, temporal and spatial horizons that exceed ordinary human capacities and institutional rhythms. Moving beyond Frankenstein Urbanism’s emphasis on rebellion and loss of control, we argue that the central problem of contemporary AI urbanism is scale asymmetry within co-constituted human and machine assemblages. The framework comprises six dimensions: cognitive scale, temporal scale, spatial scale, governance asymmetry, autonomy and agency and human response. It synthesises scholarship on AI urbanism, urban political economy, infrastructural urbanism, posthumanism, law and scale, supported by AI governance frameworks and vignettes on Hangzhou City Brain, robotaxis and generative design. The article conceptualises the urban as a constitutive condition of AI scale asymmetry, where algorithmic systems encounter infrastructure, land, mobility, public services, municipal authority, social difference and collective life. Across these domains, the key challenge is how to govern systems whose capacities outpace human oversight while remaining embedded in corporate, legal and political arrangements. We argue that effective responses require mixed-intelligence governance and temporal dignity: arrangements that subject algorithmic operations to democratic oversight, contestation and human-paced deliberation. Gulliver Urbanism explains how AI reconstitutes urban life by reorganising infrastructures, municipal authority, spatial relations, design imaginaries, public services and everyday citizenship.