AI as a Tool for Untangling Urban Complexities

Abstract

Cities are among the most complex creations of humanity and urban planning has been utilizing digital toolsets to deal with complexities of urban life for decades. Until now, most digital tools had to resort to single domain and/or declarative/deterministic approaches in analyzing urban systems and were not able to grasp a more integrated view. With the advent of more sophisticated (connectionist) AI-based toolsets, we can dive deeper into untangling the complicated, surprising, and often contradictory factors that constitute urban systems. By utilizing approaches such as GANs, computers might become able to mimic creative problem-solving and visioneering capabilities of urban planners. However, despite the hype surrounding new AI- based tools, they are often more of an improvement of existing technologies instead of substantial technological innovations, and thus need to be assessed by planners critically. The paper sets the usage of novel AI tools into the context of urban planning and discusses how urban planners and designers can interact with emerging AI technologies e.g., through Prompt Engineering. It presents a framework for the coordination and orchestration of human and artificial creativity in the processes of investigating and planning complex urban systems.