Extractions: Data Infrastructures and the Public Good

Abstract

This special issue invites reflection on digital forms of resource extractivism, through thinking-with microbes, fish, smart energy, and human bodies. We examine the confluence of infrastructure building, public good enactment, and nature datafication to account for novel forms of resource economies. The articles in this collection empirically investigate the various ways in which nature is turned into resource in laboratory and experimental sites including database retrievals and data mining. Our empirical cases, ranging from biodiversity prospecting, clinical biobanking to the digitalized electric grid and from digital fish to biodigital transformations in yeast technology, offer insights into the nexus of data, economy, nature, and information. Contributing to the literature on datafication of natural resources, these studies re-situate information infrastructures and biodigital extractions as part of emerging info-economies.

Why a Special Issue on extractions? Through thinking-with microbes, fish, smart and renewable energy, and human bodies, this collection invites reflection on how information infrastructures and data practices enable novel forms of resource extractivism. Biodiversity databases, human biobanks, or smart grids are instances of large infrastructural projects projected as key for up-and-coming info-economies. Such infrastructural projects are often supported by public funding, promoted by national states, and justified in terms of progress and the common good. We look at the confluence of infrastructure building, public good enactment, and nature datafication to account for novel forms of resource economies that draw on data mining, dissection, classification, storage, and management. Through open science, innovation policies, and large public investments in data infrastructures, commons are turned into commodities appealing to a public good that exists mainly in the form of promise. Promised goods are presented as if they were realities: in-the-pipeline data and technologies that are about to materialize or are being transformed along the way, contributing to “smartness mandates” (Halpern et al. 2017). This special issue of Science, Technology, & Human Values (ST&HV) digs into the promises and visions of a public good that is entangled in data-centric extractivism.

The various contributions to this issue offer insights on multiple ways in which nature is turned into resource through extractivist practices, in laboratories and experimental sites including database retrievals and data mining. By focusing on practices of extraction, and re-situating such practices to consider them as part of larger resource-information-public assemblages, this ST&HV thematic collection on “Extractions: Data Infrastructures and the Public Good” adds to the larger corpus of literature on extractivism. We have learned from the literature of political ecology, human geography, and anthropology how fishing, agriculture, mining, and other extractive activities have produced environmental erosion, often also resulting in conflict and dispossession (Sawyer and Gomez 2012; Cormack and Kurewa 2018; Dunlap and Jakobsen 2020). Projects for metal, gas, and oil extraction have led to waste and permanent pollution (Tironi 2018) and generated local forms of “infrastructural violence” (Appel 2012). More recently, green projects such as the construction of wind and solar plants and the local conflicts they generate have received much scholarly attention (Sovacool 2021; Harnesk and Jakobsen 2023). The kind of “infrastructural enchantment” (Harvey and Knox 2012) and promise that comes with the new green deal has been contested, this literature has recently shown. The focus in this literature has been most often on particular sites of extraction and on what infrastructural work does within those sites, including the kind of “sacrifice zones” (Reinert 2018) it generates.

The work assembled in this thematic collection approaches extractions as data practice and is concerned with processes of “data formation” (Nadim 2021) in which contexts are left behind and dismissed. For instance, laboratory work to turn specimens such as microbes, fish, or human bodies into data produces a sort of de-contextualization by which information is presented as de-linked from its original referent (Latour 1999). In this process, data and data recombinability are attributed value in their own right. The contributions to this special issue describe such processes of datafication of natural resources, attending to the kinds of de-contextualizations and recontextualizations that occur with infrastructural translations, as they are moved from nature to data and databases. Materials are informed (Barry 2005) and information becomes a good in itself. A recent corpus of literature dedicated to data extractivism (Sadowsky 2019; Smith and Shade 2018) has described such processes of information commodification, emphasizing how the publics may contest an economy in which they are both data sources and consumers in dataveillance regimes (Degli 2014). While acknowledging this potential, our conceptualization of extraction highlights the material and natural resources–oriented dimension of data-intensive economies.

This special issue opens for investigation some of those instances of datafication processes of “nature,” through empirical studies of how the ongoing reassembling of info-natures during datafication occurs. These analyses of data extractivism describe the metabolic relations that those emergent configurations enact. Cases range from value creation through biological materials to regimes of engagement in datafied energy living labs and from digital economies of metabolic modeling to multispecies encounters with the digital. Studies in this collection also attend to metabolic relations that may differ from those analyzed in the context of resource extractivism. At the same time, these multispecies assemblages often result in and enact different versions of “anthropo-econo-centrism” (Schrader 2012). Contributions to the special issue address these bio-digital and info-economic tensions through empirical cases on biodiversity prospecting (Delgado 2021), biobanking (Aarden 2022), modeling for a digital fish (Bauer 2024), biodigital yeast technologies (Szymanski 2021), and extractions through the digitalized electricity grid (Rommetveit et al. 2021).

In her contribution to this collection, Delgado looks at a turn in bioprospecting practice from taxonomic and chemical-based screenings to sequence-based bioprospecting. The paper features bioprospecting as an extractivist activity and follows the infrastructural work that goes into turning microbes from specimens to data in the search for novel and commercially interesting forms of biochemistry. The paper provides an account of extractivist practices along the biodiscovery pipeline from extraction sites in arctic and subarctic waters to large global databases. It shows how the turn to sequence information has important political implications. As sequencing and data mining technology are increasingly used to determine the commercial value of biodiversity, and databases gain importance as sites of biodiversity extraction, regulatory frameworks go out of date and biodiversity (as data) becomes an elusive public good. This paper showcases how promise is attributed to data-centric methods for exploring natural resources and for turning commons into commodities.

Rommetveit, Ballo, and Sareen (2021) also trace a shift, this time from analogue electricity extraction and distribution to the implementation of the smart digitalized grid in Norway. The paper offers an historical account of a transition from an energy economy in which hydroelectric resources were exploited and distributed locally, to a scaling up and integration of the grid at a national and later Nordic level, also enabling renewables to enter the grid. The paper shows how the implementation of the smart grid builds on power concentration and redistribution within “epistemic networks” that include industry and state actors and that relies on a representation of citizens as “active users.” In the transition toward a data-intensive electricity market, not only energy but also information on citizens’ energy habits is made into a resource for the development of new businesses and services. Smart meters, databases, and apps also construct the user as a site of extraction. Users produce data that are “granular and dynamic” and that feed back into a central database that can be used for modeling and managing domestic energy behaviors. The projection of users as active co-producers comes along with increased levels of automation and what may actually be more passive roles for users and publics. Still, the implementation of smart meters is promoted both by industry and public authorities as good for everyone because it will enable more efficiency and fair costs—a development that is presented as both socially desirable and unavoidable.

Aarden’s (this issue) article on biobanking in Singapore delves into the ways in which human biological samples attain scientific and economic value. Treating biomedical samples as resources, biobanks are key cases of promissory infrastructures, as they collect and centrally store samples linked with digital information. Their mode of managing and redistributing extracted biological materials raises questions of ownership, ethics, politics of health services, and recruitment of publics, scientific value, and population representativity. The Singapore Tissue Network, a centralized storage facility for biological samples funded by national research agencies, was expected to generate economic value for Singapore. Invested with high expectations, this biobank was shut down after a few years of operation, as it failed to yield expected economic value, both from the repository itself and from new collaborations. Aarden introduces the concept of “value formation” to capture the intersections between different forms of value in life sciences’ infrastructures. The paper conceptualizes extractions in terms of the process of retrieving value and the production of public and commercial goods out of samples and data. The study demonstrates that biobanking’s features are situated in local policies of regulation and in societal institutions that shape how decisions over profitability are made.

Bauer’s contribution to the special issue revolves around biodigital practices within Norway’s bioeconomy aspirations, where salmon aquaculture is center stage. It analyzes how a domesticated fish—the Norwegian farmed salmon—is turned into data. Empirically, the paper follows extractions from salmon liver cell cultures in the lab to the places where extracted data are brought to inform digital platforms and into modeling aimed to inform policies. As part of publicly funded infrastructure building, researchers work on a digital model of salmon metabolism, envisioned as a theory-informed prediction machine for data-driven optimizations of food, feed, and health more broadly. These models simultaneously work on optimizing benefits for human health and on improving the environmental sustainability in salmon aquaculture through reengineering soy-based feed to include more locally sourced ingredients. Interestingly, modeling metabolic relations for salmon has drawn on biomedical research infrastructures as templates for information integration. In turn, scientists also conduct optimization research on evaluating fish diets to enhance cognitive performance among publics in schools and kindergartens as interventions for better public health. While examining proliferating efforts toward datafication, the article also attends to disturbances to both aquaculture and the digital models built to inform the governance of feed, food, and health. Bauer shows how optimization not only brings marine commons into new chains of extraction and valorization but also resets the conditions for nutrition recommendations in the name of the welfare state, guided by marketable health benefits.

In the closing contribution to this ST&HV special issue on Extractions, Szymanski turns to yeast as an organism widely used in food technologies and relatively open to experimentation. She argues for maintaining connections and multiplying the practices of engagement with multispecies entities. Using “snowflake yeast”—a form of yeast growing in clusters—as a point of departure, the article examines yeast technologies as sites of biodigital transformation. The study shows how, in the case of yeast, certain digital practices do not extract in the classic sense of taking away. Instead of subtracting, by creating copies, they multiply. Hence, biology’s yeast extractions—its genome sequence—live on in online spaces, but they can also be put back again to experimentation with yeast. When it comes to laboratories working with yeast, the paper argues to keep open the relations and interdependencies that may be different from those found for natural resource extractivism. Its descriptions aim to position microorganisms like yeast simultaneously as organism-technologies and creatures with lives of their own. Rather than subsuming biodigital practices with microorganisms under extractivism, this paper calls for empirical openness when investigating digital bio-extractions and close attention to what they enact.

Together, these papers bring forward the STS literature on datafication of resources, and of data as resource, through close examination of the nexus of data, economy, nature, and information.