Art, data models and the complex city

Invited Talk at the Governing AI and Cities/Visualising AI and Cities Data Think – Winter School – Digital Visual Studies (DVS) University of Zurich – 4 March 2025 Instituto de Arquitectura Avanzada de Catalunya (IaaC) – Barcelona- Spain

Summary: The Anthropocene is not only the proposal of a new geological era characterised by ecological and geopolitical upheavals, but it is also linked to epistemological and ontological changes that have affected not only the aesthesis – the way we perceive our environment -but also the aesthetics, bringing up new values and procedures that affect the way we produce knowledge and our moral attitudes.
Linked to these changes is a new vision of the world that evolved through new sociotechnical systems. From the first image of the Earth globe taken by the astronauts of the Apollo Mission to the modelling of climate change, the Earth has appeared as a complexity of interlinked systems working at different scales. Since Global Systems can not be studied by experimental methods, and there are no data without models, the knowledge of our environment is the product of the evolution of sensing devices and data models. In other words, borrowed from K. Hayles, our knowledge is produced inside assemblages of technical and biological entities.
Cities are part of these global systems, considered the privileged human habitat and the product of human interactions and culture. At the dawn of this century, the availability of GPS systems and the spread of social networks allowed city planners to consider human beings as sensors producing real-time data. In the hands of artists, these data produced the maps of a lived city in opposition to a functional city planned around the centres of power. In his turn, in the hands of city planners, data models were applied to map the habitability of cities and improve their infrastructures. Despite the productivity of this approach, nowadays new questions arise. Lefebvre described urbanisation as a globalising phenomenon, already in the seventies of the previous century. However, he also pointed to the local specificity of the urbs, determined by its geography, climate and history, its study being a multidisciplinary task concerning social, scientific and humanistic disciplines and in which the progression moment, the proposal of what the urbs can be and its planning, must be preceded by a regressive moment, the description of what it is and the analysis of how it has become. Furthermore, new proposals for a more-than-human city are emerging, which consider habitability as more than a human question: sustainability, air quality, availability of fresh water and food, or the enjoyment of inspiring landscapes being the product of multispecies conviviality. These considerations aim to bring new perspectives to the study of the city and the usage of other data sources and models.
The multiplication of sensing devices and initiatives as open data science has made massive quantities of historical and present information affordable in multiple formats like images, historical blueprints and maps, narrative accounts, ecological inventories, and physical and chemical measurements. In front of the mentioned new standpoints for the study of the urbs and the available information raises questions: Which data models can we use to convey knowledge from this possible archive? Under what conditions does this information become actionable data? What problems arise in the application of these models?
Artists use speculatively supervised and unsupervised Machine Learning Models to make imperceptible phenomena experienceable and anticipate possible futures. My proposal for this roundtable is to see some of these artistic projects to ignite a discussion about the application of these models to the study of the urban.