In my work, I argue that smart city technologies and governance approaches can allow for the emergence of large-scale collaborative agents to stop anthropogenic environmental degradation and regenerate Earth. I view climate change as a group of problems taking on the form of coordination problems (following Guala 2016 and Chwe 2001), common-pool resource problems (following Ostrom 1990), and (super) wicked problems (following Rittel and Webber 1973 and Levin et al. 2012). While these are heterogenous perspectives, I argue that the common knowledge condition is a key feature shared by all of them. The satisfaction of the common knowledge condition (e.g., Gilbert 1989) is integral to creating group agents that can handle these types of problems, and I argue smart cities are one way to satisfy this condition. If this is in fact the case, smart cities can become group agents. Then, the question is if their agency is robust enough, and their constituent parts integrated enough, for these group agents to respond to the challenge of climate change. Thus, I create a taxonomy of smart cities as group agents: 1) the smart city, 2) the sentient city (following Crang and Graham 2007 and Shepard 2011), 3) the agentive city, and 4) the superorganismic city (following Kesebir 2012 and Wilson et al. 2013). At each level I assess the possible group agent in terms of agential capacity, based on an approach to a generalizable agential architecture (cf. Tomasello 2022). While the creation of this taxonomy is conceptually interesting in its own right, I intend to use it to clarify the potential and limitations of smart cities projects with regard to climate change. I apply each level to different use-cases, types of problems, and scales, showing that each level may be helpful for the challenges of today and tomorrow.
