AI NATIONALISM(S): GLOBAL INDUSTRIAL POLICY APPROACHES TO AI
In this essay collection we survey the nationalist narratives and emergent industrial policies being proposed by governments with differing economic and geopolitical motivations.
Overview
Collapse
In this essay collection we survey the nationalist narratives and emergent industrial policies being proposed by governments with differing economic and geopolitical motivations. The national policies surveyed in this collection have largely functioned to reinforce the notion that AI is both a socially and geopolitically important sector, and therefore worthy of government strategies (and spending) to promote it. With such far-reaching consequences for the public at stake – from the allocation of public funds to the rapid promotion of AI tools in sensitive social domains through procurement mandates – any claims to advancing public good must be scrutinized.
Why AI? Critiquing AI Industrial Policy
In the past year and a half, with the release of sophisticated AI tools to the public, interest in the speculative promise of AI technologies has exploded across the globe. AI is now firmly positioned as a critical strategic technology for the geopolitical and economic ambitions of nation-states. As governments, and the public, solidify their orientation toward the AI industry, we are forced to wrestle with the relative advantages and disadvantages of AI industrial policy — a term referring to government spending, investment, and regulatory strategies focused on the AI industry. Many governments are increasingly focused on promoting, nurturing, and growing national AI economies — and the industries that underpin them.
This uptick in government support opens up space to question: why AI? Current industrial policy frequently assumes a world in which government spending — whether on the procurement of AI products or development of the industry — is a goal that should take precedence over others in the name of ‘innovation’.
But before investing deeper in the development of an AI industry we need concrete and material answers to questions like: do efficiencies gained through AI-based climate modeling justify the energy cost of training these models? Should we invest in the advancement of edtech at the cost of providing more students school lunch, or AI medical software over funding home healthcare?
In this essay collection we survey the nationalist narratives and emergent industrial policies being proposed by governments with differing economic and geopolitical motivations. The national policies surveyed in this collection have largely functioned to reinforce the notion that AI is both a socially and geopolitically important sector, and therefore worthy of government strategies (and spending) to promote it. We find that, true to historical tradition, governments are using industrial policy as a tool to increase their own geopolitical leverage and economic competitiveness – even as they cloak those objectives underneath thinly defined “AI for social good” aims.
Perspectives in this collection also warn that industrial policy as it is currently structured functions only to further expand and concentrate private (largely US-based) power, under the banner of democratization. Simply diversifying the range of actors involved in AI development while commercial entities continue to define the horizon for development does little to contest their dominance. With such far-reaching consequences for the public at stake – from the allocation of public funds to the rapid promotion of AI tools in sensitive social domains through procurement mandates – any claims to advancing public good must be put under the scanner.
Here we build on scholarship (here and here) calling for a more democratic practice of industrial policy to reject the notion that the current trajectory of AI-centered development is inevitable.
Instead of naturalizing the idea that larger and larger scale AI is a self-evident public good, we must start by seeking a clear-eyed understanding of the ways that AI acts on our core social and economic institutions, and to whom AI’s benefits and harms accrue.
We need to look at which business models make social benefit more–or less–likely, and what the impact is on workers, the environment, and democracy. The evidence suggests that while benefits accrue to a handful of corporate actors, current AI industrial policy perpetuates a long cycle of racialized disenfranchisement of groups that reap few of the benefits and bear most of the harms. Instead, we must recognize that benefits to corporate actors often manifest as harms to the people subject to corporate AI systems.
This is why we need to look beyond our current echo chambers. For a conversation about AI and the public good to meaningfully take place, industrial policy will need to answer to the imperatives of a public beyond the tech and defense industries, and particularly to those structurally disadvantaged groups that have already borne the brunt of the costs of this industry. While this collection grapples with the limitations of the current approach to AI industrial policy, and the narratives that support it, the work ahead is a much more difficult task: articulating the kind of economy we want in the first place, and where–if at all–AI could play a part. We intend for this to serve as a provocation for taking this conversation forward, beyond the geographic boundaries of the United States and EU, and beyond the silos of tech policy debates.
Table of Contents
Collapse
See Also
Contributors
Collapse
Alejandro Calcaño Bertorelli
Administration (Do Not Display)
Max von Thun
Administration (Do Not Display)
Jyoti Panday
Administration (Do Not Display)
Mila T Samdub
Administration (Do Not Display)
Matt Davies
Administration (Do Not Display)
Sandra Makumbirofa
Administration (Do Not Display)
Islam Al Khatib
Administration (Do Not Display)
Acknowledgments
Collapse
This collection of essays was curated by Amba Kak, with editorial contributions from Sarah Myers West, and operational support from Alejandro Calcaño.
We are thankful to the authors of the essays in this collection for their generative collaboration.
Special thanks to Rishi Bharwani, Cynthia Conti-Cook, Andrea Dehlendorf, Amy Kapczynski, Julius Krein, Barry Lynn, Saule Omarova, Ganesh Sitaraman, and Meredith Whittaker for their insights and feedback on the introductory chapter and overall framing for this project.
Copyediting by Caren Litherland.
Design by Partner & Partners.
Cite as Amba Kak and Sarah Myers West, eds., “AI Nationalism(s): Global Industrial Policy Approaches to AI”, AI Now Institute, March 2024