I presented the following video at AI for Cities Workshop in AAAI conference, Austin Texas 2015
As a follow up to this fast presentation, later in collaboration with my friend and colleague Diana Alvarez-Marin, we developed several ideas and finally wrote it down with the following title:
Diana Alvarez-Marin, Vahid Moosavi, Inverting Normative City Theories and Computational Urban Models: Towards a Coexistence With Urban Data Streams, (Under review at the journal of City)
In this work, which is currently under review at the journal of City we examine the inter-relationships between what we call as computational capacities and the formation of urban theories. There, we claim that while the majority of urban modeling theories are in a way, placed in the so called tri-partite models (e.g. Kevin Lynch’s: City of Faith, City as a Machine and Ecological City or Cedric Price’s metaphoric egg models), the recent developments in computational capacities have opened up a new level (to a degree a paradigm) of urban modeling that is essentially a way out of the three previous categories.
Further, we discuss that while we study different theories from different time periods and different centuries, all of them co-exist at the current time. Therefore, we claim that these categories are not necessarily correlating with time.
The paper is a relatively long, but personally like its conclusions very much:
“We perceive currently a cohabitation of models and theories in the contemporary debate around cities and information, which reflect an interesting overlapping of the levels of abstraction we have described throughout this paper. On one side we recognize entering approaches, namely with the emerging fields of Social Physics, which focus on predictive and computational theories of human behaviour (Pentland, 2014), and of city science (Bettencourt et al., 2007). The latter has been attracting a community of physicists with a classical Newtonian approach, asserting the possibility of an absolute universal principle for cities or any kind of human agglomeration, able to prescribe energy consumption and size, while decoupling from any form of cultural contextualisation. This approach asserts consequently that cities are purely definable in terms of fixed universal laws, stressing therefore deterministic ideas of the city and the human, where natural conditions override cultural ones.
Concurrently, on the second level, which we identify as knowing, pre-defined algorithms become primary over data. Mostly well known and merchandised under the larger flag of “Smart City”, these models are based on the transposition of enterprise managerial processes onto the urban realm and embody private economic interests. Supporters of this approach see integrated digital infrastructures for collecting and processing data as a second electrification for cities. Defining cities merely in terms of their infrastructural makeup, may it be physical or digital; this approach presents as ultimate purpose the optimization and performance of the city as finite and controllable urban object. More computationally hungry and complex than the entering approaches, and its social physics models, machinic models are composed by groupings of rules, translated into systems of algorithms or machines. Oppositely to the previous level of abstraction, these rules are not natural or divine, but dictated by administrative and economic powers (interests).
Alongside, we identify a third approach or bottom-up view of the Smart City, led by hackers, activists and startups. We will relate these views to a connecting approach to the city, as networks and relations between agents or machines become primary, and disaggregated models overcome the limits of previously centralized ones. In this line of thought, as Ratti and Townsend put it on the “The Social Nexus” (Ratti and Townsend, 2011)., the best way to harness a city’s potential for creativity and innovation is to jack people into the network and get out of the way. Inspired from Jane Jacobs view of city neighbourhoods as organs of self-governance as described in “The Death and Life of Great American Cities” (Jacobs, 1961), this view pre-assumes horizontal structures in city planning would work better than vertical structures. During the last years, this approach has emerged more concretely with the development of civic apps in the early 2010’s and continues growing today with the rapid development of digital economy through platforms such as Airbnb, Uber, and crowd funding services such as Kickstater and Indiego. However and paradoxically, in spite of promising a more productive way forward by integrating formerly ignored citizen needs and promoting entrepreneurship, these platforms rely on some kind of centralised corporate states, which provide an infrastructural setup for connectivity. In addition to underlying socio-economical unevenness of its data sets, the connecting approach takes perhaps too jauntily for granted the idea that social networks are fundamentally democratic.
Technology is the answer but what was the question? It’s been almost fifty years since Cedric Price provocatively titled his 1966 lecture with a question that comes back today with tremendous actuality, at a moment where to more complex situations we keep on offering more complex solutions, bigger systems, more parameters, more agents, more computational power. However, Cities have never been about efficiency or optimization. Certainly, 11000 years of urban history barely match with the current standards promoted by rankings of sustainability and liveability. As Greenfield (2013) points out, at the moment, we are only being offered one particular story about the deployment of networked informatics in the urban milieu, and though it is widely predominant in the culture, it only portrays the narrowest sliver of what is possible. Today, with the advent of an increasingly important digital urban make up, we believe architects and urban researchers should look with genuine interest and curiosity at transversal developments in computation and computer science.
This is the reason why we have attempted through our urban genealogy, by taking a step back and looking more abstractly, at illustrating how concurrent informational and technological concepts have a direct influence in the construction of imaginaries and theories about the city. The relation between information technologies and cities has existed since their very inception. However, with the fast development of computers and the exponential increase in computational power followed by the data deluge, we identify an epistemic gap deepening between the level of abstraction these information technologies allow for and our models and theories about the city. When we consider machines purely as fast calculators and confer them with power through simplistically assembled statistics, we are reducing complex entities such as cities to the same kind of deterministic scope, be it centralized or decentralized.
As we have shown on this paper, there’s a cohabitation of bodies of thought or epistemes in the contemporary urban realm, yet their scope seems to limit to previous technological capacities. People behave in unexpected and irrational ways that can in no way be predicted by any kind of underlying system, and so are cities, full of paradoxes and contingent occurrences. We believe the crucial question is not the axiomatic solution of problems by establishing a brand data-driven science or theory, nor oppositely the upcoming of a brand new empiricism. While there are new instances that don’t seem to fit into the current urban discourse, we see the unfolding of a fourth level of abstraction, which is about learning and tends towards a focus in the constitution of problems. “A Quantum City, Mastering the Generic” (Hovestadt et al., 2015) embodies a data intensive or indexical idea of the city, which we consider is along this line of thought. Under this light, the perception of patterns should not be considered as answer but rather the key towards new set of questions. This paper was written as an introduction to our body of research, and in our future works we will aim at elucidating the new possibilities that this fourth level on learning is opening up.”