233046 SE Data that Matter(s): From Promises to Practices to Residues (2026S)

Our everyday understandings and discourses of ‘the digital’ are replete with concepts, metaphors and other imagery of immateriality. ‘Cyberspace’, ‘virtuality’ or ‘the cloud’ – such notions suggest that ‘online’ is less material than ‘offline’, that virtual reality is independent from physical reality, that ‘the digital’ somehow exists on a different ontological plane (Furlano, 2019). This premise of the immateriality of data continues to inform much of the research and development, as well as contemporary policy discourses on the economic and environmental values of data.

Research within Science and Technology Studies (STS) and adjacent fields, most notably Critical Data Studies, has challenged these conventional ways of speaking and thinking about data. It shows how the dichotomy of virtual/physical invisibilize what is the material center of ‘the digital’: undersea cables, data centers and databases, raw materials and mining practices, as well as the labor of factory workers, clickworkers and informalized e-waste scavengers (among many other things). If “‘raw data’ is an oxymoron” (Gittelman, 2013) and data is always already “cooked” (Bowker, 2013), we should ask: Who are the cooks, where are the kitchens and how do we cook them?

The seminar “Data That Matter(s)” invites students with an interest in critical STS perspectives on digitalization, data and artificial intelligence. It will introduce them to the different ways STS(-inspired) research has investigated the ways in which data come(s) to matter. Together, we will explore the heuristic value of what we might call a material turn in the critical study of data and AI. Through collective readings and discussion of select scholarly contributions, the seminar will provide students with new ways of making sense of the socio-material “hinterlands” (Law, 2004) of contemporary debates about data and AI. Through keeping data diaries, students will learn to reflect on their entanglements with the materialities of the allegedly immaterial digital to ponder ways of making them more care-ful and response-able.

We consider ‘mattering’ (e.g., Butler, 1993) to be a heuristic that invites us to think both about the performative effects of data and AI as they fundamentally transform everyday practices and the multiple materialities that constitute the situated, socio-material assemblages hidden behind and constituting ‘AI’ and ‘data’. We will follow the mattering of data and AI along three lines that will structure our seminar:
1) imaginaries and promises: Data and AI come with immense promises of economic growth, enhanced knowledge production, geographic flexibility and environmental sustainability. Far from being pipe dreams, these imaginaries and promises for the future have performative and material effects in the present. They inform policymaking, funding and investments as well as societal discourses.
2) practices: Not only do data and AI transform everyday practices, accustomed ways of doing things, but they are also the product of situated and idiosyncratic practicesoften living under labels such as data compilation, curation and cleaning. Data work involves heterogeneous practices of sense-making and interpretation that consequently shape how we know with/through data and AI.
3) residues: Data and AI pile up different kinds of material left-behinds throughout their existences that are, to a large extent, unacknowledged and difficult to assess and care for. Servers on which data are stored require raw materials, extracted through mining, which leave behind toxic wastes. The “exhaust” (Zuboff, 2019) of our digital practices creates massive amounts of unused data – data waste(s) (Felt, 2020) – stored in data centers that gobble immense amounts of energy and water. Finally, components of digital technologies will be turned into heaps of e-waste, often piling up in informal dumps in the Global South.