Farming Data, Forming the Cloud: The Environmental Impact and Cultural Production of Information Technology
Farming Data, Forming the Cloud:
The Environmental Impact and Cultural Production of Information Technology
Asta Vonderau, Department of Social Anthropology, Stockholm University
Facebook’s decision to build a mega data center in Luleå in 2013 in order to provide server cooling and storage facilities for the social network’s vast amount of data marked the start of a major economic development: the global cloud infrastructure’s migration towards the Northern hemisphere. This opened up new possibilities for Sweden to attract IT industries’ investitions, and to generate innovative ideas for regional development. Accordingly, the new industry was welcomed by business developers, based on popular, dematerialized images of the cloud and on the IT industry’s self-promotion as the cheap and green 21st century antidote to a toxic past centuries’ factory work. Cloud industries’ and infrastructures’ enormous energy and water needs, their underinvestigated long-term environmental effects (waste heat), and the authorities’ limited possibilities to legally control global IT companies’ activities were seldomly discussed.
This project aimed to investigate the local social and environmental effects of cloud infrastructure and related industries based on ethnographic research at Facebook’s Luleå location. It addressed the internet’s materiality and the risk of an energy crisis of information. The project asked: How is the ‘cloud’ imagined? How does the ‘cloud’ materialize in terms of environmental change? And how is the ‘cloud’ socially negotiated in terms of social relations and the labor market?
The project developed largely according to plan. During the empirical research phase, the focus of the study has been broadened beyond the exclusive concentration on Facebook and its data center in Luleå in order to include studies of Luleå’s and Norrbotten’s emerging (post-)industrial ecosystem, and of transnational processes of expert knowledge transfer. This change was partly due to the expected high securitization and difficult accessibility of Facebook which necessitated complementary entry points to the research field from the start. It also quickly became clear that a focus on one single IT company would be misleading, given my aim to study the sociocultural effects of cloud infrastructure as relational socio-material configurations, thus linking multiple localized technologies, expert groups, places and institutions to each other.
As originally planned, the first project year was devoted to preparing and conducting multi-sited ethnographic research. Starting with a case study of Facebook’s data center project, the fieldwork also included an investigation of transnational and global dimensions of cloud infrastructure and industry. It was conducted at different sites that included Stockholm, Monaco and Hannover among others. The fieldwork resulted in a solid base of empirical materials, consisting of interviews; ethnographic observations; and a collection of political documents, archival materials, and media and administrative reports. During the second and third project year, collaborations with other scholars were established, and the project’s results widely disseminated. I have made more than 20 project presentations for international academic and non-academic audiences and organized one anthropological and one interdisciplinary international workshop in Stockholm. Project results contribute to current debates in Anthropology, but also Media Studies, Human Geography, Environmental Studies, and other disciplines which aim at a critical and environmentally reflexive investigation of the consequences of digitization.
The projets most important results include the following:
1. Generating knowledge about the internet’s materialities and their local entanglements. While popular media discourses about IT are still dominated by dematerialized images of the internet (and the cloud) as a virtual, global and fluid entity, researchers now acknowledge the need to investigate its materialities. Qualitative studies at IT infrastructural sites are, however, still very rare. My project filled this gap by generating knowledge about how exactly global cloud infrastructure is entangled in concrete social and political contexts. It shows how IT giants such as Facebook strive to reduce their local industrial and infrastructural visibility in order to downplay the cloud’s existing resource needs, and to motivate limitless internet use. The Swedish state, in turn, promotes Facebook’s local presence in order to attract more IT infrastructure providers, and to brand Sweden as a world leading IT-nation. At the same time, regional authorities support such industrial development as they aim to empower the region vis-á-vis the national centres, in the hope of turning Norrbotten into an independent IT competence hub. The mutual dynamics between these actors and their interests triggers both planned and unexpected effects: National territories are rescaled; natural landscapes change; new modes of governance and new subjectivities (as for instance new expert elites) emerge.
2. Understanding the geopolitical logics of digital capitalism. My ethnographic study opens up possibilities for relating and comparing different IT infrastructural locations. It allows to map how this industry’s effects are distributed throughout global geographies of the cloud. It demonstrates that major IT companies shape the cloud as a translocal “technological zone” (Barry 2006) which is supposed to enable frictionless flows of capital and data. Such economic interests are not always in line with the aspirations of national or regional actors. Accordingly, the emerging glocal geographies of the cloud which connect Luleå and Sweden to other places not only shape new centres and competence hubs, but also new peripheries to which resource extraction and environmental effects of IT networks and services are outsourced. Bearing in mind the data center industry’s environmental impacts and limited local job opportunities, it is of major importance to investigate the border-crossing social inequalities between smart centers and infrastructural peripheries of digital capitalism.
3. Contributing to the methodological toolbox of anthropological (and interdisciplinary) infrastructure studies. During recent years, anthropologists have made efforts to develop new methodological tools for investigating complex socio-technological assemblages such as infrastructures. These scholars acknowledge the changing relations between state and private economy in a globalized and connected world, and aim to develop synchronic perspectives on different forms of sociality without making a priori distinctions between global forces and local places, human and non-human actors, culture and nature. This project has contributed to infrastructure studies’ toolbox: it provides examples on how to get access to IT infrastructures’ secretive and invisible dimensions by means of ‘polimorphous engagements’ (Gusterson 1998, Vonderau 2018); it developps a relational perspective on the cloud which allows to investigate how infrastructural scales and images of ‘the global’ are made relevant and real (Vonderau 2018); and it provides tools for the investigation of the ‘poetics of infrastructure’ (Larkin 2013), that is, the ways how infrastructures disconnect from their technical functions and integrate in people’s life worlds.
In the course of this project it became obvious that in practice, the national authorities’ possibilities to legally regulate global IT companies on their territory remain limited and unclear. It is important to investigate these difficulties and their consequences in concrete settings, and to collect evidence about if, and how, the new EU General Data Protection Regulation may change this relation between state and IT business or not.
To calculate the data center industry’s environmental impact is not easy. Little information is disclosed by the companies, and their impacts are distributed unequally within potentially global geographies of the cloud. Hence, it is necessary to investigate these inequalities. The best way to conduct such an investigation is collaborative research involving social, IT, and environmental scholars.
The public perception of the internet and the cloud is often still dominated by commercial and dematerialized images which obscure materialities and potentially problematic aspects of digitization. It is necessary to continue conducting research which makes the internet’s materialities visible and shows how they shape life-worlds and natural environments.
IT infrastructural developments and their social and environmental impacts are border-crossing processes which relate different places and groups of people to each other. Accordingly, the project’s fieldwork was multi-sited, mobile, and international. Ethnographic research on cloud infrastructure is still rare. Apart from this project, there is no such research conducted in Sweden. Most relevant studies on this topic have been conducted by anthropologists and media scholars in the UK and the US. Therefore, networking and collaborating with international scholars has been essentially important throughout the project.
Project results were presented in different countries including Sweden, UK, Germany, Italy, and Lithuania, for academic and non-academic audiences, as for instance industry experts at The Node Pole, SICS ICE data center initiative, and Data Center Dynamics. Project publications were published in English and German. As project leader and solve principal investigator I have organized two international workshops in Stockholm: Infrastructuring: Materialities of Time, Power and (In)justice (October 2017, together with Mark Graham) and Investigating Data Centers: Socio-Technical Assemblages of the Cloud (November 2017). These workshops gathered international scholars and artists working on the topics of the cloud and digital infrastructures. The second workshop was organized in collaboration with Bahnhof Pionen data center. It resulted in an additional publication, a special issue of the peer-reviewed journal Culture Machine.
I also was invited to become a guest researcher at HafenCity University Hamburg (Januar-March 2017) where I presented my research and developed further research ideas on IT infrastructures in relation to smart city projects together with scholar from CityScience lab. Between October 2017 and July 2018, I even acted as a visiting professor at the Department of Cultural Anthropology, Goethe University Frankfurt Main. Here, I taught in the new international and interdisciplinary master study program, Science and Technology Studies: Economies Governance, Life. My teaching included seminars on cloud and IT infrastructures, thus internationally disseminating project results.
Project results also were published in form of several academic articles in both disciplinary as well as interdisciplinary journals and popular science publications. This publishing strategy proved to be the best way to disseminate results to different academic and non-academic audiences. The majority of the project publications are published open access: Imaginations, Culture Machine and Zeithistorische Forschungen are open access journals. The journal Ethnos publication is also planned to be open access. Also, Standpunkte (the publication of Rosa-Luxemburg Foundation) is open access. Only the Zeitschrift für Volkskunde does not offer open access, mirroring the general situation in Germany where possibilities for open access publications still are not very well developed.