Data centres form an integral part of current global information infrastructures. The data centre is where 'the cloud' becomes experiential. The panel addresses novel methodologies for researching infrastructures and materialities from within STS, anthropology, sociology and media studies.
Data centres form an integral part of current global information infrastructures. The data centre is where data lives, where 'the cloud' becomes somewhat experiential, where scholars may expect to get a handle on the mattering and worlding of data. Nevertheless, at data centres we are confronted with assortments of components, visions and systems that appear incongruous and, at times, trite: corporate strategies of liquid workforces, integration, security; grey corridors, grey panels, black metal cages; gas cylinders, diesel engines, water pipes, cables. Through the work of Susan Leigh Star we know that the invisibility and boredom of(in) data infrastructures are essential to their functioning as infrastructure; and that it takes careful attention and analysis to bring into focus the politics of(in) data infrastructures. While Star's work remains extremely pertinent, the importance of (big) data has been proliferating in the last years. The specificities of data distribution, storage and maintenance - where and when it happens, how it is done and by whom - are shot through with histories and politics. The life cycle of data is structured around myriads of decisions, taken in the name of efficiency, security or practicability, that reproduce differences and, more often than not, inequalities.
The panel addresses site-specific collaborations at data centres. We are particularly interested in papers that engage with the temporality, spatiality and modularity of data, as well as invite contributors to discuss methodologies for researching infrastructures and materialities, data engagements or the folding of data from within STS, anthropology, sociology and media studies.
This panel is closed to new paper proposals.
Locating the data centre. Attending to place in the study of cloud computing
This paper will examine cloud computing as a localized practice that has a place - the data centre. We will describe the relations between distributed data flows and the sturdy materiality of place.
This paper explores ways of locating cloud computing in the material infrastructures of cloud storage. Attending to place when it comes to delocalized, distributed data storage can help understand the ways in which practices with data are involved in infrastructuring and making worlds. Experimenting with research into digital infrastructures in Upper Franconia (Germany), we will ask for the close ties of cloud infrastructures with their surroundings, temporally, spatially and in terms of regional economies. In doing so, we will focus on the specific segmentations on site: We will address the spaces inside the data centre, as well as the connectivities between inside and outside in order to understand the modularity, spatial and temporal relations within extended data machines. In the data centre itself, we find infrastructures of storage embedded in broader business strategies of liquid workforce, as well as tedious air-conditioned rooms filled with stacks of hardware and the monotony of server sound. In its attempt to locate, track, and follow such connections, this talk will investigate the containment and flow of data as well as their foldings that reach beyond the data centre. We will discuss the socio-geographical and historical relations at the site, the repurposing of architecture, the generation of data security via redundancy at multiple levels as well as constantly adapted business models of future value creation. Combining approaches from STS and media studies, we invite a discussion on methodologies for collaborative, interdisciplinary research into the materiality of digital infrastructure.
Prospecting the domains: data, knowledge, software and much else
"Domain" is a keyword in contemporary data science ventures and a long standing keyword for the computing and information sciences. "Domain" serves to identify and bound a target for prospecting (or farming), and serves to organize relations between expert participants.
Farming data is today's tip of the historical iceberg in efforts to prospect the domains. "Domain" is an actors' category in the computing, information and data sciences -- it refers to any worldly expertise, for instance, biology as the "domain science of life" or geology as the "domain science of the earth". The term "domain" serves to identify the target of efforts in the computing sciences to bring within their purview some feature of that domain: today data, but at other times, knowledge, software, workflows, ontologies and so on. This presentation will place today's efforts to farm data within a longer historical trajectory of prospecting the domains, and argue that increasingly the logic of domains has become the de facto organizing principle for today's computing sciences, encoded in the organization of data centers, in toolsets and in funding policy.
The more GPUs you buy, the more money you save: images in data centres
The paper will explore how large accumulations of images affect the architectural dynamics of data centres. It points to the importance of images as operational elements of data centres as technical ensembles.
This paper presents an attempt to speak appropriately about the movement of images into data centre operations and platform infrastructures. The paper will explore the mixture of mundane combinations and unpredictable transformations that have appeared in some spectacular platform demonstrations such as Facebook's 'planetary settlement model' from the standpoint of the GPUs, or graphics processing units, that have become central to many predictive models working with large scale image collections in data centres. It traces the embedding of large image collections in a GPU-based model as a process that transforms collections of images into a highly condensed indexical field, a field that allows platforms to generate predictive propositions or statements about the world. In describing how predictive indexicality takes shape on platforms and machine learning models, the altered collectivity of images under data-centre platform conditions becomes more evident. This indexicality also suggests the need to map data centres as techno-affective architectures in a flow of images.
Preparing for future value - collaboration in big data labs
Mushrooming big data labs are sites of intense collaboration towards new data habitats. This talk explores how preparing for big data means re-configuring the data centre along with shared imagined futures and responsibilities.
My notebook and I are standing in an insurance company's big data lab. A poster near me offers an insight into the ongoing collaborations right now. It depicts grey shaped icons and arrows in one direction, attached to it, coloured post its with handwritten notes. The data centre is in flux and demands collaboration of different actors, including posters and post-its, but also the re-organization of responsibilities and data itself.
In my ethnographic field stay I observed how different actors engaged in preparing for big data. This involved re-configuring the data centre, the data stream, the data itself and people's relations to each other and the data. This re-configuration draws on existing infrastructures of responsibility, data work and valuation of data. In this paper I show that big data labs are sites of infrastructuring where several actors, technologies and visions come together in order to collaborate for a new data infrastructure, including a new data habitat. This results in productive frictions between various actors, practices and concepts, i.e. between infrastructure and security experts.
Throughout the nine months of fieldwork, the envisioned data centre fulfilled a variety of functions to various actors, from being home of new imagined (big) data value to being a flexible object of collaboration. The big data lab is where the new data infrastructure comes into being while re-configuring relations of people, technologies and economic value. .
Data meets electronic waste: practices and temporalities of a transformative encounter
This talk focuses on the intersection of data and electronic waste handling and highlights data centres as potential hubs of waste accumulation. The data-electronic-waste handling at our university serves as the empirical starting point to dive into the manifold transformations of data_carriers.
There is much concern about data - in public and in scholarly discussions. Less in the spotlight are data carriers, the very basic condition to bring data into existence. But what happens if data or its carrier is considered to be ‚of no use' any longer? What happens when data becomes dispensable and/or the functioning of the data carrier is impaired? Recently, STS researchers have been examining data centres as a means to grasp the slippery tangibility of data and to ground its socio-material constitution. In ecological terms, data centres are commonly framed as energy-wasting facilities. Yet, in this talk we shift the gaze towards another, more neglected waste-problem of data centres: the intertwined existence of electronic and data waste. Engaging with practices and temporalities at the intersection of data and electronic waste, we will highlight data centres as potential hubs of waste accumulation. Here, waste of different kinds comes into being, is handled, and processed. Focusing on the politics and practices of waste in data infrastructures, we approach the death and/or afterlife of data and its carriers. The data-electronic-waste handling at our university serves as the empirical starting point to dive into the manifold transformations of data_carriers: from data to data garbage to e-waste to secondary resources. Distributed responsibilities combined with the simultaneity of trust and ignorance are at heart of negotiations and practices. Caring for data while neglecting e-waste and vice versa is enacted by particular actors at specific times of transformation.
The cloud at the horizon of the internet infrastructure
The cloud and the trend to cloud computing is changing internet infrastructure and has also effects on data infrastructures. This paper aims at giving answers to the questions that arise thereof, including centralisation and power relations.
The development of the "cloud" has become one of the most important trends of changing IT and data infrastructures and it is still gaining speed. Although most of the developments subsumed under cloud computing are technologically and economically driven, the cloud metaphor and the cloud's inherent changes to the material and technological infrastructures are not neutral. DeNardi's and Musiani's 'governance by infrastructure' assumes purposefully created infrastructure in order to regulate and govern via this technology, i.e. 'frozen governance'. I add to this concept the analysis of diffuse structural and co-constitutive power relations, also engaging with the methodological consequences that this entails. I argue that the trend of cloud computing is characterising a further centralisation of the internet with effects on the control of functionality and data, challenging traditional ideas of government/governance. A core question to be addressed is: how does "the" cloud change power relations with regard to democratic institutions, structures, businesses and society? Trends and changes that can be summarised under the umbrella of increasing cloud usage are related to a "reassortment" of (constitutive) social and power relations between actors, technology and values. Cloud computing leads to the creation of intelligent (data) centres where the computing and storage takes place. Clients such as phones and computers become rather passive displays of the information rendered by the network. This altering of the internet's traditional and emancipatory end-to-end principle is more than a technological change, but creates new dependencies and changes power relations both within and outside of the network.
Promissory collaborations: Big-Tech datacenters and the state
This paper analyses the emergence of the datacenter industry in Denmark as a site of promissory collaboration between the state and Big-Tech, arguing that datacenters provide an interesting perspective from which to think about transformations in data politics.
Big-Tech's arrival to Denmark - the datacenters of Apple, Facebook, and Google - has excited imaginaries of prosperity, progress and regional transformation. Such globally expansive tech corporations are now entering into promissory forms of collaboration with the state; allegedly providing what states are increasingly finding it difficult to offer on their own; jobs, infrastructure and viable futures.
Through a preliminary examination of Big-Tech in Denmark, his paper analyses the materialities and politics of locating data and digital infrastructures in small energy-rich nation states. With a critical eye towards such techno-political interventions, and an analytical sensitivity to the geographies (where) and temporalities (when) of data, I suggest that the move towards globally distributed, yet locally situated, digital infrastructures sheds light on how data, and its allegedly transformative potentials, are becoming ever more embedded into the modes by which other relationships form and unfold.
I do this by analyzing how datacenters in Denmark are not only sites of assemblage - gathering the photos, music and personal data of billions of dispersed users - but are also sites of promissory collaboration that offer a particular perspective on the ways in which entities such as the state and Big-Tech corporations are under constant reformulation in a rapidly transforming dataverse. At the same time, paying attention to the where and when of data in distributed infrastructures prompts me to provide a lateral reflection on the where and when of data in ethnographic practice.
This panel is closed to new paper proposals.