Using multilayered individual massive data and geo-localization tracking, Computational Social Science and Social Physics attest the importance of face-to-face and place-to-place interactions in shaping human agency. The research suggests that collective and individual behaviors can be described by the “social bridges” that bind local communities in daily activities. By looking at various metropolitan regions across the world, findings point to the resilience of social bridges in predicting economic, political and health-related characteristics of local populations. Communities that share similar practices of metropolitan spaces also express similarities in those behaviors. Such results are akin to theories of urbanism that promote the importance of spatial configurations in describing markets, power and well-being. With this paper, I wish to look at the empirical findings of Computational Social Sciences and Social Physics through the lens of those theories, and propose a general understanding of human agency based on aspirations, affordances and capabilities. Instead of competing with previous approaches, the view that individuals act in accordance to their environment as an extension of their desires and aptitudes provides a long-waited hinge to bind Social Physics and previous theoretical models. In a second step, I discuss how such model can afford tools to architects, planners and policy-makers to increase collaboration and cooperation between local communities and decrease the economic and political polarization of contemporary metropolitan spaces.