Artificial intelligence as planetary assemblages of coloniality: The new power architecture driving a tiered global data economy

Abstract

We present a framework for viewing artificial intelligence (AI) as planetary assemblages of coloniality that reproduce dependencies in how it co-constitutes and structures a tiered global data economy. We use assemblage thinking to map the coloniality of power to demonstrate how AI stratifies across knowledge, geographies, and bodies to influence development and economic trajectories, impact workers, reframe domestic industrial policies, and reconfigure the international political economy. Our post-colonial framework unpacks AI through its (1) global, (2) meso, and (3) local layers, and further dissects how these layers are vertically integrated, each with its horizontal dependencies. At (1) the global layer of international political economy maps a new digital bipolarity expressing Sino and American global digital corporations’ strategic and dominant positions in shaping a tiered global data economy. Then, at (2) the meso layer, we have a mosaic of domestic industrial policies that fund, frame markets, and develop AI talent across industries, sectors, and organizations to competitively integrate into AI value chains. Finally, incorporating into these are (3) the localized labor processes and tasks, where workers and users enact various AI-mediated tasks and practices driving further value extraction. We traced how AI is an interlaced system of power that reshapes knowledge, geographies, and bodies into dependencies that reinforce stratifications in developing underdevelopment. This commentary maps the current digital realities by laying out an uneven techno-geoeconomic power architecture driving a tiered global data economy and opening new research avenues to examine AI as planetary assemblages of coloniality.

Keywords

Artificial intelligence tiered global data economy planetary assemblages AI value chains dependency international political economy

This article is a part of special theme on Datafied Development. To see a full list of all articles in this special theme, please click here: https://journals.sagepub.com/page/bds/datafied_development?pbEditor=true

Introduction

We present a post-colonial framework that sees artificial intelligence (AI) as planetary assemblages of coloniality that are driving a tiered global data economy. This framework builds on the “decolonial turn” in critical data studies (see Casilli, 2017; Couldry and Mejias, 2023), to help us see how AI reproduces dependencies by co-constituting and structuring an uneven power architecture driving a tiered global data economy. Data have long been used to sort, order, and categorize in ways that uphold social and power asymmetries. This is further reinforced at scale by using AI to process massive amounts of data. Using assemblage thinking (Hansen and Thylstrup, 2024; Kitchin, 2022) to map out key aspects of the coloniality of power, which illuminates the post-colonial structures of control and hegemony reproduced as modernity and in rationality, and how they affect diverse spheres of life—such as our ways of being, knowing, and feeling (Quijano, 2000, 2007; Ricaurte, 2019), we aim to unfold how AI shifts power and influences development trajectories by reshaping knowledge, geographies, and bodies into dependencies that reinforce stratifications in developing underdevelopment (Cieslik and Margócsy, 2022; Frank, 2000; Treacy, 2022).

At the time of writing this piece, a tenth of the global economy is digital. Of the top one hundred global digital corporations worth a total of US$15 trillion in market capitalization, the Americans and Chinese account for 90% of this valuation (Hosseini, 2023; Steinhoff, 2023).¹ Underlying their valuations are these global digital corporations’ investment, control, and ownership of data and associated infrastructures (Narayan, 2022). Based on this, the rest of the world only accounts for about < 10% of the global data economy. These developments suggest the hyper-concentration of value capture in making a tiered global data economy skewed towards the elites in the United States or China. Their value-capture processes operate across scales with interconnecting layers, renewing post-colonial power relations and dependencies that connect the peripheries in this new Majority World to their Sino and American dominant cores. These dynamics continue the ongoing enclosure and erasure of other ways of knowing and subjectivities in and from the peripheries to privilege data-driven rationality as the dominant epistemology. As a result, AI's development reproduces an uneven power architecture, driving a tiered global data economy reflecting the coloniality of power.

Our post-colonial framework unpacks AI as planetary assemblages of coloniality through its specific global, meso, and local layers (see Figure 1). Then, it dissects how these interconnected layers are vertically integrated, each with its complementing horizontal dependencies. At its core, this framework traces how AI and its data flows are entangled in and are the product of an uneven distribution of power and resources at various levels. It emphasizes the skewed political economy of knowledge production, building out technical divisions of knowledge, expertise, and skills that privilege dominant data epistemologies as exemplified by connectionist forms of AI (see Bender et al., 2021; LeCun et al., 2015). These technical divisions progress into social and international divisions across geographies and bodies as classes, races, and genders, worsening stratifications as new sites of value and planetary extraction. As sites of extraction, they are part of an emerging techno-geoeconomic power architecture articulating AI as planetary assemblages of coloniality. This new power architecture serves the strategic interests of elites situated in digital bipolar cores, impacting workers, reframing meso or domestic industrial policies, and reconfiguring the international political economy.

Figure 1.

Artificial intelligence (AI) as planetary assemblages of coloniality with layered horizontal dependencies.

Viewing AI beyond just technologies and techniques and toward planetary assemblages of coloniality

Data processing using AI has become the most advanced way of representing knowledge as reflected in their increased use-value in terms of expertise and skills. When we see AI as technologies and techniques, they are composed of amalgams of beliefs, practices, methodologies, instruments, and infrastructures by which the data lifecycle is enacted based on advanced statistics and stochastic models. For workers and students, these technical amalgams are made seductive and aspirational as vehicles to advance or gain employment and access social, economic, and cultural capital. We see this in how AI-related skills augment workers’ average hourly wage by over 20%, where the demand for machine learning and the use of Google Brain's TensorFlow can command as much as a 40% increase (Stephany and Teutloff, 2024). These AI-related in-demand knowledge, expertise, and skills convey their economic function and valuation as commodities that can be voluntarily rented out to interested employers. However, this economist-functionalist view of AI remains too constrained to fully explain how these technical amalgams, which were once at the fringes of the AI discipline in the 1980–1990s, are part of dominant discourses of development (see World Bank, 2021), economic growth, and aspirations at present.

The re-emergence of AI within a longer history of techno-economic changes cannot be explained by the technologies and techniques alone (Acemoglu and Johnson, 2023; Perez, 2010). The changing fashions of which knowledge, expertise, and skills are in demand suggest that the current AI-mediated work practices, requirements, and epistemologies are neither neutral nor operate in a vacuum. Framing AI merely as an amalgam of technologies and techniques erases its political nature by obscuring the interests and the promoters that it serves (Suchman, 2023). This framing also eliminates questions of class and class conflicts that intersect with post-colonial, solidarity, and planetary struggles across stratified knowledge, geographies, and bodies. Instead, we should regard the current techno-economic changes as existing within wider social processes imbued with the politics of coloniality, where AI is not a singular object but is composed of “a multitude of interlaced systems of power” (Crawford, 2021: 12; Tacheva and Ramasubramanian, 2023). Therefore, we contend that planetary assemblages of coloniality are particularly useful in tracing how AI as the dominant epistemology are not just technical amalgams but are ensembles and products of specific histories, material infrastructures, social processes, and prevailing politics that span across the planet in outlining the contours of our current techno-capitalist realities and plausible futures.

Our framework: Three vertically integrated layers of AI

In our recasting of AI as planetary assemblages of coloniality, this framework examines its three key vertically integrated layers of the local, their in-between meso, and their ties to their global layers. As a planetary assemblage of coloniality, AI is crafted to control and capture digital value by integrating across these key interconnecting vertical layers (see Figure 1). At the top, (1) the global layer of the international political economy maps a new asymmetric digital bipolarity, expressing the strategic and dominant positions of the Sino and American global digital corporations in shaping a tiered global data economy. Then, (2) at the meso layer of national and subnational domestic industrial policies, fund, frame markets, and develop AI talents across industries, sectors, and organizations. These industrial policies, together with skills development, talent systems, labor, and visa regimes, aim to competitively integrate domestic economies and cross-border trade into various AI value chains. Finally, incorporating into these meso-level policies and AI talent pipelines are the (3) localized labor processes within organizations at the points of production, where workers and users enact various AI-mediated tasks and practices driving further data-value extraction. 1.

The global layers’ horizontal dependencies driving a tiered global data economy.

As planetary assemblages of coloniality, AI is composed of interconnecting layers reinforcing relations of horizontal dependencies. First, we zoom out to the top of our assemblage to see the global layer of AI as a new international political economy reflecting the return of great power competition. From this international political economy vantage point, the story of how the new global power architecture is regionalizing globalization is a bipolar geoeconomic one expressed through asymmetric Sino and American techno-corporate powers as duopolistic empires with overlapping security, trade, and cultural interests forming their coalitions (Lee, 2018; Mirrlees, 2024). Articulating this new asymmetric digital bipolarity, global digital corporations, such as Microsoft, Apple, , NVIDIA, Alphabet (Google), Meta (Facebook)and Amazon (MANAMA), and others in the United States exert significant influence over the global data economy due to their pioneering role in capturing about 70% of market value. Baidu, Bytedance, Alibaba, Tencent, Xiaomi (BBATX), and others from China collectively hold 20%. These trends are also exhibited in AI-related knowledge production (Koch et al., 2021). Competition is nothing new, especially with the return of para-state companies and their imperial logic,² raising concerns of geoeconomic fragmentation and uncertainties ahead (Butollo, 2021; Tan and Yang, 2021). We find indications of this in the World Trade Organization's regionalized arrangements in rules governing digital trade and e-commerce (Ismail, 2023), displaying blocs or coalitions of distrust and new partnerships in moving us into a multipolar world.

With the United States and China accounting for 90% and the rest of the world at < 10%, this presents a “tiered global data economy [that] is set to become the norm,” where a few geographies, organizations, and actors have cornered a disproportionate share of the gains (IT for Change, 2020: 3). Their top AI and technology firms have become de facto “global digital corporations with planetary reach” by their capture, control, and investments in all parts of AI value chains (UNCTAD, 2019, 2021: xvi). These Sino and American global digital corporations perform a vital role in assembling the global power architecture by building the physical scaffolding of network infrastructures and connectivity, likened to the building and control of strategic colonial ports and railways of the past. Moreover, their capture, control, and investment in data, software, hardware, standards, and infrastructures, such as undersea cables, satellites, and data centers, as the new governing stack architecture that connects all corners of the periphery to the US or China establish structural dependencies through lock-ins and bottlenecks across the globe (Bratton, 2016; Lehdonvirta et al., 2023). Within their inter-imperialist Sino and American elites’ rent-seeking value networks acting as coalitions, these cores assert structural control over critical infrastructures and resources, especially AI talent, compute power (Sastry et al., 2024), and data to assert their world-making visions as dominant epistemologies over their dependent geographies and bodies. 2.

The meso layers’ horizontal dependencies through industrial policies.

To make this international political economy vantage point more concrete, we move from the top layer toward the in-between meso layer of AI within an American-centric coalition by presenting the broad trends of Canada and India’s³ domestic industrial and employment policies on data and AI. These industrial policies are experimenting with re-industrialization through greening digitalization in Canada, India, and elsewhere like the European Union in the rest of the world as new digital peripheries seek to competitively integrate into Sino or American global AI value chains (Butollo et al., 2022; Howson et al., 2021; Sturgeon, 2021). Both Canada and India, as semi-peripheries in a tiered global data economy, are strategically reorganizing their production and work systems across sectors and within organizations to enhance their domestic capabilities. Firms and higher educational and national AI institutions play key roles in screening and developing AI talent pipelines for their domestic ecosystems in defining their labor and visa regimes. Canada's industrial policies focus on the higher-end value activities of model-related research, talent development, and commercialization (ISED, 2022). As part of Canada's integration into global AI value chains, sufficient workforce development in AI and upskilling professionals are pillars of its industrial policy to generate an estimated 16,000 jobs by 2028 along the Québec-Waterloo corridor (Scale AI, 2020).

India adopted a more mechanistic strategy to foster their National AI Marketplace (NAIM) with the “AI value chain” in mind, with an emphasis on lower-value activities like data-related work, including annotation (MeitY, 2024; NITI Aayog, 2018: 78). Despite India producing 2.6 million STEM graduates in 2016 alone, they mainly work in IT services, and not research and development (MeitY, 2024; NITI Aayog, 2018: 78). Most of its graduate researchers migrate to the United States, the United Kingdom, or Canada, with India reportedly retaining less than 400 of its 22,000 PhDs (Chahal et al., 2021; NITI Aayog, 2018). As a distinction, India plans to reposition and strategically integrate into global AI value chains as a global “AI Garage for 40% of the world,” where enterprises and organizations can incubate, develop, and refine data solutions before rolling them out in other Global South sites (Nandi and Yadav, 2024; NITI Aayog, 2018: 19). Canada, on the other hand, has positioned itself by focusing more on AI research and model work, with a concentration on commercialization with its North Atlantic allies. From this, we see the relative positions of Canada and India along global AI value chains, where India is looking at the Majority World and focusing more on data work, whereas Canada is focusing on the wealthier Minority World in doing more work on AI models and research upstream. As illustrative cases, the respective positions of Canada and India as digital semi-peripheries make clearer the horizontal dependencies at the meso or domestic layer, owing to their relative specializations and geographies, segmenting AI-related value-added activities within an American-centric coalition. 3.

The local layers’ horizontal dependencies between workers and users.

Finally, moving further down to the local layer, we have workers and users who are a part of AI's development and labor processes. At this layer, the tiered global data economy makes the stratifications and the international divisions of labor more distinct. Local variations and experiences with regulating these dependencies remain more contingent and mutable as the global and meso layers act more subtly in the background in our day-to-day lives due to their specific labor and visa regimes. As the dominant trend, these stratifications again build out from the technical into their social and international divisions of labor in AI development in bifurcating work, workers, and their bodies. Sambasivan et al. (2021) shed light on this trend where “data is often the least incentivized aspect, viewed as ‘operational’ relative to the lionized work of building novel models and algorithms.” As a result, “under-valuing of data work is common to all of AI development … where the Global South is viewed as a site for low-level data annotation work, an emerging market for extraction from ‘bottom billion’ data subjects, or a beneficiary of AI for social good.”

Most of the tedious data work for AI production is outsourced to the peripheries or semi-peripheries like India, while higher-valued model work remains closer to techno-geoeconomic cores, such as Canada in its relative proximity to the United States. At one end are the countless hidden data workers in India, for instance, doing the data work, which is the process of preparing and curating the data tasks necessary to build the databases and to train machine learning models (Gray and Suri, 2019; Hung, 2024; Hung et al., 2023; Le Ludec et al., 2023; Miceli and Posada, 2022; Muldoon et al., 2024; Tubaro et al., 2020). On the other end, the highly visible and prized model workers are doing model work, which is the building of computational programs or algorithms to generate models. Model workers, such as AI engineers, researchers, and data scientists, are normally seen in male-dominated and higher-paid professions. Pockets of model workers and data workers also exist within Canada and India's local layers. However, the international divisions of labor detailed beforehand remain the foremost tendency. At this embedded local layer, the technical, social, and international divisions in the work of making AI further segment laboring bodies with classes, races, and genders as systems of violence and dispossession become clearer (Couldry and Mejias, 2023; Gray, 2023; Muldoon et al., 2024; Thatcher et al., 2016). By viewing AI as planetary assemblages of coloniality, our post-colonial framework presents the broad contours of how AI develops and reinforces stratifications in developing underdevelopment by structuring layered horizontal dependencies.

Conclusion

Viewing AI as planetary assemblages of coloniality that reinforce extractive relations of a dependency across scales opens us to examine how it reproduces an uneven power architecture driving a tiered global data economy. We traced how AI is an interlaced system of power that is vertically integrated across its global, meso, and local layers, each with their respective horizontal dependencies, by reshaping certain knowledge, geographies, and bodies in a way that reinforce stratifications in developing underdevelopment. We showed an uneven power architecture that benefits a select few and affects workers, domestic industrial and employment policies, and the international political economy. As this emerging power architecture divides us into a tiered global data economy that mainly serves the elites in the United States or China, we must rigorously interrogate and transform the politics of coloniality found in data-driven rationality. This type of framework provides researchers with new avenues to study AI relationally across its vertically integrated layers to grasp its extractive horizontal dependencies. For policy and practice, including regulations and contestations, it provides human heuristics by outlining present digital trajectories to inform change strategies and acts of epistemic disobedience across scales and layers to decolonize and re-affirm our own ways of being, knowing, and feeling. Future collective regulations, good governance, and research on making AI more responsible must address its uneven power architecture and skewed political economy in driving a tiered global data economy.

Footnotes

Acknowledgments

The author thanks Christian Lévesque, Cassandra Bowkett, Fabian Ferrari, Susan Spronk, and Phoebe Moore for their support and editorial input, as well as the anonymous reviewers and the editors' helpful comments.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the International Development Research Centre (IDRC) Research Award Recipient, Interuniversity Research Centre on Globalization and Work (CRIMT) Studentship, Social Sciences, and Humanities Research Council (SSHRC) Canada Graduate Scholarship.

ORCID iD

Kai-Hsin Hung

Notes

References

Acemoglu

Johnson

(2023) Power and Progress: Our Thousand-Year Struggle over Technology and Prosperity, First edition New York: PublicAffairs.

Bender

Gebru

McMillan-Major

, et al. (2021) On the dangers of stochastic parrots: Can language models be too big? FAccT 2021 - Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency: 610–623.

Bratton

(2016) The Stack: On Software and Sovereignty. Cambridge, MA: MIT Press.

Butollo

(2021) Digitalization and the geographies of production: Towards reshoring or global fragmentation? Competition and Change 25(2): 259–278.

Butollo

Gereffi

Yang

, et al. (2022) Digital transformation and value chains: Introduction. Global Networks 22(4): 585–594.

Casilli

(2017) Digital labor studies go global: Toward a digital decolonial turn. International Journal of Communication 11: 3934–3954.

Chahal

Abdulla

Murdick

, et al. (2021) Mapping India’s AI Potential. CSET Data Brief. Center for Security and Emerging Technology. Available at: https://cset.georgetown.edu/publication/mapping-indias-ai-potential/.

Cieslik

Margócsy

(2022) Datafication, power and control in development: A historical perspective on the perils and longevity of data. Progress in Development Studies 22(4): 352–373.

Couldry

Mejias

(2023) The decolonial turn in data and technology research: What is at stake and where is it heading? Information, Communication & Society 26(4): 786–802.

10.

Crawford

(2021) Atlas of AI: Power, Politics, and the Planetary Cost of Artificial Intelligence. New Haven, CT: Yale University Press.

11.

Frank

(2000) The development of underdevelopment. In: Sociological Worlds. Routledge, pp. 86–93.

12.

Gray

(2023) More than extraction: Rethinking data’s colonial political economy. International Political Sociology 17(2): olad007.

13.

Gray

Suri

(2019) Ghost Work: How to Stop Silicon Valley from Building a New Global Underclass. Houghton Mifflin Harcourt.

14.

Hansen

Thylstrup

(2024) Stack bricolage and infrastructural impermanence in financial machine-learning modelling. Journal of Cultural Economy 17(1): 20–38.

15.

Hosseini

(2023) Plattformökonomie 2023: Amerika dominiert, EU fällt zurück. Available at: https://www.platformeconomy.com/blog/newtop100.

16.

Howson

Ferrari

Ustek-Spilda

, et al. (2021) Driving the digital value network: Economic geographies of global platform capitalism. Global Networks 22(4): 631–648.

17.

Hung

K-H

(2024) A Case of Assessing the Working and Living Conditions of Data Workers in India's Global Artificial Intelligence Value Chains. Ottawa: IDRC Digital Library. https://hdl.handle.net/10625/63473.

18.

Hung

K-H

Bowkett

Garneau

, et al. (2023) Algorithmic management and workspace power dynamics: Redefining the frontiers of visible and invisible work. In: Les Intelligences Artificielles Au Prisme de La Justice Sociale. Considering Artificial Intelligence Through the Lens of Social Justice. Quebec, Les Presses de l’Université de Laval, pp. 299–320. Available at: https://www.degruyter.com/document/doi/10.1515/9782766301850-013/html.

19.

ISED (2022) Government of Canada launches second phase of the Pan-Canadian Artificial Intelligence Strategy.

20.

Ismail

(2023) The Evolving Context and Dynamics of the WTO Joint Initiative on E-commerce: The Fifth-Year Stocktake and Prospects for 2023 . International Institute for Sustainable Development. Available at: https://www.iisd.org/system/files/2023-04/wto-joint-initiative-e-commerce-fifth-year-stocktake-en.pdf .

21.

IT for Change (2020) Unskewing the Data Value Chain: A Policy Research Agenda for Equitable Platform Economies . Available at: https://itforchange.net/sites/default/files/1805/IT%20for%20Change%20-%20Concept%20note%20-%20Unskewing%20the%20data%20value%20chain.pdf.

22.

Kitchin

(2022) The Data Revolution: A Critical Analysis of Big Data, Open Data & Data Infrastructures. 2nd edition. London: Sage Publications Ltd.

23.

Koch

Denton

Hanna

, et al. (2021) Reduced, reused and recycled: The life of a dataset in machine learning research. In: Proceedings of the Neural Information Processing Systems Track on Datasets and Benchmarks 1 (NeurIPS Datasets and Benchmarks 2021) round2. NeurIPS.

24.

LeCun

Bengio

Hinton

(2015) Deep learning. Nature 521(7553): 436–444.

25.

Lee

K-F

(2018) AI Superpowers: China, Silicon Valley, and the New World Order. Harper Business.

26.

Lehdonvirta

Hawkins

(2023) Cloud empires’ physical footprint: How trade and security politics shape the global expansion of U.S. and Chinese data centre infrastructures. SSRN. Available at: https://papers.ssrn.com/abstract=4670764.

27.

Le Ludec

Cornet

Casilli

(2023) The problem with annotation. Human labour and outsourcing between France and Madagascar. Big Data & Society 10(2): 20539517231188723.

28.

MeitY (2024) Cabinet Approves Over Rs 10,300 Crore for India AI Mission, will Empower AI Startups and Expand Compute Infrastructure Access . Ministry of Electronics & IT. Available at: https://pib.gov.in/PressReleaseIframePage.aspx?PRID=2012375.

29.

Miceli

Posada

(2022) The data-production dispositif. In: Proceedings of the ACM on Human-Computer Interaction 6 (CSCW2). Article: 460. Association for Computing Machinery, pp. 1–37.

30.

Mirrlees

(2024) The US and China’s digital tech war: A new rivalry within and beyond the US empire? Socialist Register 60(A New Global Geometry?): 105–140.

31.

Muldoon

Cant

, et al. (2024) A typology of artificial intelligence data work. Big Data & Society 11(1): 20539517241232632.

32.

Muldoon

Graham

Cant

(2024) Feeding the Machine: The Hidden Human Labour Powering AI. London: Canongate Books.

33.

Nandi

Yadav

(2024) Digital Dreams, Real Challenges: Key Factors Driving India’s AI Ecosystem. RF Occasional Paper No. 436, May 2024, Observer Research Foundation. Available at: https://www.orfonline.org/research/digital-dreams-real-challenges-key-factors-driving-indias-ai-ecosystem.

34.

Narayan

(2022) Platform capitalism and cloud infrastructure: Theorizing a hyper-scalable computing regime. Environment and Planning A: Economy and Space 54(5): 911–929.

35.

NITI Aayog (2018) National Strategy for Artificial Intelligence Discussion Paper . Available at: https://www.niti.gov.in/sites/default/files/2023-03/National-Strategy-for-Artificial-Intelligence.pdf.

36.

Perez

(2010) Technological revolutions and techno-economic paradigms. Cambridge Journal of Economics 34(1): 185–202.

37.

Quijano

(2000) Coloniality of power, eurocentrism, and Latin America. Nepantla: Views from South 1(3): 533–580.

38.

Quijano

(2007) Coloniality and modernity/rationality. Cultural Studies 21(2–3): 168–178.

39.

Ricaurte

(2019) Data epistemologies, the coloniality of power, and resistance. Television and New Media 20(4): 350–365.

40.

Sambasivan

Kapania

Highfill

, et al. (2021) “Everyone wants to do the model work, not the data work”: Data cascades in high-stakes AI. In: Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems. Yokohama Japan: ACM, 1–15, 6 May 2021. Available at: https://dl.acm.org/doi/10.1145/3411764.3445518.

41.

Sastry

Heim

Belfield

, et al. (2024) Computing Power and the Governance of Artificial Intelligence. arXiv:2402.08797. arXiv. Available at: http://arxiv.org/abs/2402.08797.

42.

Scale AI (2020) Strategic Plan 2020. Montreal, QC. Available at: https://cdn.scaleai.ca/uploads/2020/05/scale-ai_strategic-plan_en.pdf.

43.

Steinhoff

(2023) The industry of automating automation: The political economy of the AI industry. In: Handbook of Critical Studies of Artificial Intelligence. Edward Elgar, pp. 312–322.

44.

Stephany

Teutloff

(2024) What is the price of a skill? The value of complementarity. Research Policy 53(1): 104898.

45.

Sturgeon

(2021) Upgrading strategies for the digital economy. Global Strategy Journal 11(1): 34–57.

46.

Suchman

(2023) The uncontroversial ‘thingness’ of AI. Big Data & Society 10(2): 20539517231206794.

47.

Tacheva

Ramasubramanian

(2023) AI empire: Unraveling the interlocking systems of oppression in generative AI’s global order. Big Data & Society 10(2): 20539517231219241.

48.

Tan

Yang

(2021) The new liability of origin in global decoupling. Management and Organization Review 17(3): 624–629.

49.

Thatcher

O’Sullivan

Mahmoudi

(2016) Data colonialism through accumulation by dispossession: New metaphors for daily data. Environment and Planning D: Society and Space 34(6): 990–1006.

50.

Treacy

(2022) Dependency theory and the critique of neodevelopmentalism in Latin America. Latin American Perspectives 49(1): 218–236.

51.

Tubaro

Casilli

Coville

(2020) The trainer, the verifier, the imitator: Three ways in which human platform workers support artificial intelligence. Big Data and Society 7(1): 1–12.

52.

UNCTAD (2019) Digital Economy Report 2019: Value Creation and Capture - Implications for Developing Countries. Geneva: United Nations Conference on Trade and Development.

53.

UNCTAD (2021) Digital Economy Report 2021 : Cross-Border Data Flows and Development : For Whom the Data Flow. Geneva: United Nations Conference on Trade and Development.

54.

World Bank (2021) World Development Report 2021: Data for Better Lives. Washington, DC: World Bank. Available at: https://openknowledge.worldbank.org/handle/10986/35218.