Leveraging Generative AI Models in Urban Science.

Area of Science:

• Urban Studies
• Computational Social Science
• Artificial Intelligence

Background:

• Cities are the dominant human habitat, necessitating understanding of urban environments' impact on behavior and perception.
• Urban form and social behavior are intricately linked, requiring advanced methods to study their interplay.
• Foundation models represent a novel AI approach for analyzing complex urban dynamics.

Purpose of the Study:

• To explore the relationship between urban form and social behavior using advanced AI.
• To simulate individual-environment interactions in 21 global cities.
• To evaluate the capacity of foundation models in capturing urban complexities and human perceptions.

Main Methods:

• Analysis of urban morphology indicators and data collection methodologies.
• Application of foundation models (a new AI generation) for simulation.
• Scrutiny of AI model performance in representing human behavior and city perception.

Main Results:

• Foundation models can simulate individual-environment interactions across diverse urban settings.
• The study highlights the potential of AI in urban science for understanding complex city dynamics.
• AI models demonstrate capacity, yet require careful evaluation for urban research applications.

Conclusions:

• Advanced AI, specifically foundation models, shows promise for urban research and policymaking.
• Meticulous evaluation of AI capabilities and limitations is crucial for effective application.
• Understanding human-AI interaction within urban contexts is key for future urban development.