In lieu of an abstract, here is a brief excerpt of the content:

Reviewed by:
  • Resisting AI: An Antifascist Approach to Artificial Intelligence by Dan McQuillan
  • Nathan Schneider
Resisting AI: An Antifascist Approach to Artificial Intelligence
by Dan McQuillan
BRISTOL UNIVERSITY PRESS, 2022, 190 PP.
PAPERBACK, $29.95
ISBN: 978-1-529-21350-8

In Cadwell Turnbull’s story “Monsters Come Howling in Their Season,” a mid-twenty-first-century artificial intelligence on the island of St. Thomas is democratically governed by its users.1 It aids them in adapting to and mitigating climate change; when it misbehaves they can elect to adjust its settings. Over the course of the story, the protagonist begins to trust the system with his own life’s data and accepts the terms of service.

This is not an embrace that scholar, technologist, and activist Dan McQuillan expects anytime soon. In his book Resisting AI: An Antifascist Approach to Artificial Intelligence, the antifascist approach to AI is in at least one respect simple: opposition.

McQuillan’s book is deliberate and generous with its readers. Chapters proceed logically and sequentially, with brief summaries at the beginning and the end of each. The author draws on complex ideas in social, political, and media theory, but never to the point of showing off. From start to finish, this brief work drives toward the abolitionist point that the only viable thing to do with AI is to resist its every incursion into our lives. Redemption is a lost cause. “Existing systems, and AI perhaps most of all,” he writes, “are not simply tools that can be turned to good ends or bad but technosocial infrastructures with an established momentum” (146).

The first and longer portion of Resisting AI consists of the critique. AI is not actually intelligent, McQuillan argues, and what it does do largely consists of reinforcing the worst inequalities and inhumanities of the present world order. It segregates populations, punishes without cause, cruelly deprives, and allows elites to evade responsibility. Drawing on Achille Mbembe’s notion of “necropolitics,” McQuillan argues that AI amplifies a system of power that “not only discriminates in allocating support for life but sanctions the operations that allow death” (85).

The second and shorter portion of the book turns to the texture of life in resistance while seeking to make good on McQuillan’s promise at the outset to deliver “an ultimately optimistic text” (1). The final chapters turn to feminist thought for a standpoint epistemology centered on the people most likely to experience harm from AI, combined with an ethic of care. He proposes nonbureaucratic popular assemblies—workers’ councils and people’s councils beyond the workplace—as a starting point for direct democracy that can set an agenda outside those of corporate and state power. These, he expects, can form the basis of a movement that enacts a radical strategy of refusal, mutual aid, and technology based on freely associating communities of peers. [End Page 211]

In the last pages, AI gets an olive branch, an invitation to join in the making of society’s transformation. “For AI to be part of this change,” McQuillan suggests, “it must replace optimization with commonization” (145). Perhaps Turnbull’s benevolent AI is possible after all. But McQuillan is reluctant to even call it AI, so different would need to be the technology’s underlying design, purpose, and role in society. He does not reveal much about what the redeemed technologies would be or do, besides the centrality of care and governance by assembly. There is an allusion by citation on the last page to Project Cybersyn, the worker-coordinating computer envisioned in Salvador Allende’s socialist government in 1970s Chile. By the end the message is clear that we should resist the new robot overlords, but the counterprogram remains more in the realm of optimism than well-grounded hope.

The track record of ad hoc, anti-institutional movement assemblies such as McQuillan calls for has been tragically meager in recent years, especially up against platform capitalism and authoritarian states. Cooperative tech platforms have floundered against the might of venture-capital-backed competitors. Peer production practices have been far more likely to be drafted into service for big tech than to...

pdf