Kate Darling · Nonfiction · 2021
Darling argues we should model our relationship with robots on our history with animals — partners and tools we neither mistake for humans nor pretend are mere machines — rather than on the android fantasy. The review takes the analogy as a useful corrective to anthropomorphic hype while flagging its limit: animals are not engineered by firms with incentives to manufacture attachment, which is exactly where AI companion design needs governance.
Human-Robot RelationsAnthropomorphismEthics
Allison J. Pugh · Nonfiction · 2024
Pugh's sociology defends "connective labor" — the interpretive, emotional work of making another person feel seen, in teaching, medicine, and care — as exactly the kind of work automation flattens when it treats connection as inefficiency. The review uses it to sharpen the site's labor thread: the question is not whether a model can simulate empathy, but what institutions lose when they accept the simulation as a substitute.
Connective LaborCare WorkAutomation
Adam Becker · Nonfiction · 2025
Becker, a physicist, dismantles the grand techno-futurism of Silicon Valley — superintelligence salvation, space colonization, immortality — as a quasi-religious "ideology of technological salvation" that launders present power into promises about the future. The review treats it as a direct counter to the site's own subject matter: when AGI is framed as destiny, the claim is rhetorical, and the real question is who gets to own the future being promised.
Tech FuturismIdeology CritiqueLongtermism
Nicholas Diakopoulos · Nonfiction · 2019
Diakopoulos surveys how algorithms already shape journalism — automated story generation, newsbots, audience targeting, and computational story-finding — and argues the goal should be hybrid systems that keep human editorial judgment in the loop. The review draws the line to generative AI in newsrooms: automation reshapes not just output volume but what counts as newsworthy and who is accountable for an error published at machine speed.
JournalismAutomationEditorial Judgment
Gerd Gigerenzer · Nonfiction · 2022
Gigerenzer argues that algorithms beat humans in "stable worlds" with fixed rules and abundant data, but falter amid genuine uncertainty where human heuristics shine — and that conflating the two is how hype outruns evidence. The review pairs this with the site's risk-literacy theme: deferring to a model is sometimes wise and sometimes a category error, and knowing which is itself a skill institutions must keep.
Risk LiteracyHeuristicsHuman Judgment
James Boyle · Nonfiction · 2024
Boyle treats personhood as a line-drawing practice — law has long decided which entities can own, sue, consent, or be harmed (corporations, animals, rivers, and, shamefully, excluded humans) — and asks what happens when fluent AI becomes a new claimant. The review keeps his discipline: the danger is not proving machines conscious but powerful actors routing accountability through artificial proxies, so governance (logs, revocation, a reachable responsible human) must come before metaphysics.
LawPersonhoodAccountability
Reich, Sahami & Weinstein · Nonfiction · 2021
A philosopher, a computer scientist, and a policy scholar argue that Big Tech's core failure is a mindset of optimization — maximizing a chosen metric while treating values and trade-offs as someone else's problem. The review reads it as a direct diagnosis of AI development: the objective function is a political choice, and "we just optimized the metric" is how consequential decisions get laundered into math that democracy never reviewed.
Tech PolicyOptimizationDemocracy
Micklitz, Pollicino, Reichman et al. (eds.) · Nonfiction · 2021
This Cambridge volume asks how constitutional ideas — rule of law, due process, fundamental rights, separation of powers — translate when public and private power is exercised through algorithms. The review treats it as the public-law companion to the site's governance pieces: AI accountability is not only an ethics or engineering problem but a constitutional one, about which guarantees survive when decisions move into automated systems.
Public LawFundamental RightsGovernance
Adrian Mackenzie · Nonfiction · 2017
Mackenzie treats machine learning as a "data practice" — a set of concrete operations (vectorizing the world, drawing decision boundaries, tuning) through which messy reality is made computable. The review uses it to puncture the magic: a model's authority comes from mundane, contestable engineering choices about representation, not from understanding, so "the model predicts" is always also "someone decided how to encode this."
Critical Data StudiesMachine LearningSoftware Studies
N. Katherine Hayles · Nonfiction · 2012
Hayles argues for "technogenesis" — humans and their technologies co-evolve, reshaping attention and cognition (her hyper- vs. close-reading distinction) as media change. The review draws the through-line to AI: tools that read, summarize, and recommend at scale do not just deliver information, they retrain how people attend and think, so the interface is a cognitive intervention before it is a convenience.
Media TheoryCognitionDigital Humanities
Lisa Gitelman (ed.) · Nonfiction · 2013
This edited volume's title is its thesis: data is never simply found, it is always collected, cleaned, framed, and interpreted by people with purposes. The review treats it as foundational for the archive's data-centric pieces — every AI training set is a manufactured artifact, so "the data shows" is a claim about choices and conventions, not a window onto unmediated fact.
Data StudiesMedia HistoryEpistemology
Phil Jones · Nonfiction · 2021
Jones examines microwork platforms — the data labeling, content moderation, and task work that train and prop up "automated" systems — and argues this labor is deliberately fragmented, underpaid, and rendered invisible. The review pairs it with the site's other labor sources: what gets sold as AI autonomy often rests on a hidden, globally dispersed workforce that the interface is designed to make you forget.
MicroworkPlatform LaborGlobal South
Paul Dourish · Nonfiction · 2017
Dourish argues that information has materialities — the specific forms of spreadsheets, databases, network protocols, and emulators shape what can be represented, computed, and known. The review pulls the thread into AI: a model's "knowledge" is constrained by the concrete formats and pipelines it is built from, so claims about machine intelligence are also claims about the unglamorous material substrate the interface hides.
MaterialityInformation InfrastructureRepresentation
Natasha Dow Schüll · Nonfiction · 2012
Schüll's ethnography of Las Vegas slot machines shows addiction engineered into the interface — the "machine zone" where time, money, and self dissolve into a tuned loop of continuous play. The review reads it as the design blueprint behind today's engagement-maximizing AI: the danger is not a conscious machine but an institution that optimizes a feedback loop against a human nervous system, then calls the result a feature.
Persuasive DesignAttentionCompulsion Loops
Mark Coeckelbergh · Nonfiction · 2020
A compact primer (MIT Press Essential Knowledge) on the core questions — responsibility, transparency, bias, privacy, and whether machines can be moral agents — written to orient non-specialists rather than settle debates. The review values it as a map of the terrain while pressing its quietest point for this archive: the moral weight of a system comes from its consequences and the humans accountable for it, not from any inner life it lacks.
AI EthicsPhilosophy of TechnologyResponsibility
Daniel Susskind · Nonfiction · 2020
Susskind argues that "task encroachment" will gradually leave many people unable to compete for paid work, and that the real challenges are then distribution, the concentrated power of those who own the machines, and the question of meaning once work no longer supplies it. The review treats his economic case as the useful core while pressing the political one it raises: who controls the surplus, and who decides what a post-work life is for.
AutomationEconomics of WorkDistribution
Sheera Frenkel & Cecilia Kang · Nonfiction · 2021
Two journalists reconstruct how Facebook's leadership repeatedly chose growth and reputation management over safety as scandals mounted. The review reads it as a case study in the "architecture of denial" — the org structures, incentives, and messaging that let a platform defer accountability — and treats it as prologue to AI governance, where the same instinct to ship first and explain later now operates at model scale.
Platform PowerAccountabilitySocial Media
Melanie Mitchell · Nonfiction · 2019
A computer scientist's clear-eyed tour of what modern AI actually does and where it breaks — the "barrier of meaning," brittle perception, and the gap between pattern-matching and understanding. The review prizes its core discipline for this archive: separating demonstrated capability from extrapolated hype, and treating "the system does not understand" as a sober engineering fact rather than a slogan.
AI LiteracyMachine LearningHype Critique
Ben Shneiderman · Nonfiction · 2022
Shneiderman rejects the framing that pits human control against machine autonomy, arguing systems can be both highly automated and highly human-controlled — reliable, safe, and trustworthy by design. The review takes his two-dimensional framework as a corrective to "human-in-the-loop" hand-waving, then asks the governance question it raises: high control only matters if the controls are legible, reachable, and answerable to the people the system acts on.
Human-Computer InteractionDesignSafety
Bruce Schneier · Nonfiction · 2023
Schneier generalizes "hacking" from computers to every rule-governed system — tax codes, markets, law, democracy — where the powerful find and exploit loopholes faster than the rules can adapt. The review extends his frame to AI: a system that can discover and execute exploits at machine speed turns hacking from a human craft into infrastructure, which makes governance, patching, and who-gets-to-hack the central questions.
SecuritySystemsPower
Susan Schneider · Nonfiction · 2019
A philosopher's caution for the AI-interface age: before anyone markets machine minds, uploads, or brain enhancements, we need a clear account of what is being claimed about persons, experience, and identity. The review reads it as a guard against the "consciousness trap" — neither inferring a subject from fluent conversation nor dismissing the question — and pairs it with the site's rule that governance (disclosure, accountability) should come before metaphysics.
Philosophy of MindConsciousnessIdentity
Martin Ford · Nonfiction · 2021
Ford's argument is that AI is a general-purpose technology on the order of electricity — a utility that will seep into everything — and that its disruption to jobs, security, and power is therefore systemic, not sector-by-sector. The review takes the "AI as utility" frame as the useful core, then asks the governance question Ford raises but cannot resolve: a utility this pervasive demands public accountability, yet it is being built and metered by a handful of private firms.
Economics of AIAutomationGovernance
Paul Daugherty & H. James Wilson · Nonfiction · 2024
The Accenture authors argue the real action is the "missing middle" — hybrid roles where people train, explain, and sustain AI while AI amplifies human work — and offer a corporate roadmap (MELDS, fusion skills) for getting there. The review takes the optimism on its own terms but flags what the playbook underweights: who funds the retraining, who absorbs the displacement, and whether "augmentation" survives contact with a quarterly cost target.
Future of WorkAugmentationBusiness Strategy
Joseph E. Aoun · Nonfiction · 2024
A university president's program for an education that machines cannot easily automate, built on "humanics" — technological, data, and human literacies plus creativity and lifelong learning. The review credits the framing but presses the harder question the slogan glides past: "robot-proof" is a moving target set by whoever builds the robots, so an education staked on out-running automation inherits the same instability it means to escape.
EducationFuture of WorkLiteracy
Frank Pasquale · Nonfiction · 2020
A legal scholar proposes four "new laws" that reframe automation around complementing rather than replacing human professionals, and around forbidding systems that counterfeit humanity or entrench unaccountable force. The review reads it as a governance counter-program to efficiency-first AI: the goal is not maximal substitution but preserving the judgment, accountability, and expertise that professions are supposed to hold in trust.
Automation PolicyProfessionsLaw
Caroline Criado Perez · Nonfiction · 2019
A documentation of the "gender data gap" — the way datasets, defaults, and designs treat the male body and male life as the norm, from crash-test dummies to drug trials. The review reads it as essential background for AI bias: a model trained on records that already under-count half the population does not invent the gap, it scales and launders it into decisions that look neutral.
Data BiasGenderRepresentation
Orit Halpern & Robert Mitchell · Nonfiction · 2023
A critique of "smartness" as a governing logic — the demand that cities, infrastructures, and populations be made sensable, optimizable, and resilient through ubiquitous computation. The review uses it to name what the site keeps circling: AI is not just tools but a planetary mandate to manage uncertainty by measuring everything, and that mandate carries its own politics about what counts and who decides.
Planetary ComputationGovernanceCritical Theory
David G. Robinson · Nonfiction · 2022
A close history of the algorithm that allocates donated kidneys, and the years of contested public deliberation that shaped it. The review holds it up as a working counter-model to the site's cautionary cases: legitimacy did not come from the math being optimal but from a slow, participatory process about whose values the formula would encode — the kind of governance most AI systems skip.
Algorithmic GovernancePublic ParticipationEthics
Brian Christian & Tom Griffiths · Nonfiction · 2016
A popular tour of how computer-science ideas — optimal stopping, explore/exploit, caching, scheduling — map onto everyday human decisions. The review takes the friendly framing seriously but pushes on its edge: once these heuristics are embedded in systems that decide for us rather than advise us, the elegant math becomes governance, and "the optimal policy" quietly encodes whose costs and whose time the objective was tuned to minimize.
Decision-MakingComputer ScienceCognition
Ben Buchanan & Andrew Imbrie · Nonfiction · 2022
Two security scholars frame AI through its "three sparks" — data, algorithms, and compute — and three temperaments: evangelists, warriors, and Cassandras. The review takes the book's central wager seriously: that AI's trajectory is a contest between democratic and authoritarian uses, and that whether the new fire warms or burns depends on governance choices, not the technology alone.
AI and GeopoliticsNational SecurityDemocracy
Hannah Fry · Nonfiction · 2018
A mathematician's tour of algorithms loose in the world — in courts, hospitals, policing, and cars — that resists both hype and panic and keeps returning to one question: when do we defer to the system, and when do we override it? The review reads its "keep a human in the loop" instinct as necessary but unfinished, since the harder problem is designing for the moment a confident algorithm is wrong and a tired human is supposed to notice.
AlgorithmsHuman JudgmentPublic Trust
Wendy Hui Kyong Chun · Nonfiction · 2021
A media theorist's argument that machine learning does not merely find patterns but manufactures them — encoding correlation, homophily, and historical segregation into proxies that then pass as neutral recognition. The review reads it as the theoretical counterweight to the site's empirical sources: bias is not a bug to be debugged out but a politics built into how these systems sort people into neighborhoods of risk and resemblance.
Algorithmic BiasMedia TheoryRecognition
Mullaney, Peters, Hicks & Philip (eds.) · Nonfiction · 2021
An essay collection insisting that computing is not weightless: it runs on minerals, water, power, and human labor, and carries the histories of race, gender, and empire in its design. The review uses it to ground the site's "material AI stack" theme — the cloud is somebody's data center, and the costs of a model are paid in places the interface never shows.
InfrastructureHistory of ComputingLabor
Ifeoma Ajunwa · Nonfiction · 2023
A legal scholar's account of how the modern workplace turns employees into streams of data — through hiring algorithms, wearables, and productivity tracking — and how employment law has lagged behind the surveillance it enables. The review reads it as the doctrinal backbone for the site's labor essays: the measured worker is governed by systems whose criteria are hidden, whose errors are costly, and whose oversight the law has only begun to build.
Worker SurveillanceEmployment LawLabor
Hilke Schellmann · Nonfiction · 2024
An investigative reporter tests the AI tools that now screen résumés, score video interviews, and monitor workers — often by feeding them her own application — and documents how opaque, weakly validated systems make consequential calls about hiring and firing. The review reads it as the workplace edge of algorithmic management: a control system whose errors are hard to see, harder to contest, and rarely answerable to the people they rank.
Workplace SurveillanceHiring AlgorithmsLabor
Ajay Agrawal, Joshua Gans & Avi Goldfarb · Nonfiction · 2022
The economists' follow-up to Prediction Machines argues that AI's real disruption is not point solutions dropped into old workflows but "system solutions" that redesign how decisions get made — and that the lag between the two is where the value, and the upheaval, actually sit. The review reads the gap as a governance window: redesigning the system also relocates who decides and who is accountable, and that choice is rarely made by the people the redesign acts on.
Economics of AIDecision-MakingAutomation
Dan Davies · Nonfiction · 2024
Davies revives Stafford Beer's cybernetics to explain the "accountability sink" — an arrangement in which a decision is diffused through process, policy, and system until no person can be held responsible for it. The review reads this as the operating manual for AI deployment: when a model sits in the loop, the sink deepens, and "the algorithm decided" becomes the most efficient way yet invented to make a choice that no one will answer for.
CyberneticsInstitutional PowerAccountability
Michael Kearns & Aaron Roth · Nonfiction · 2019
A computer scientist's case that values like privacy and fairness can be written directly into algorithms — through tools like differential privacy and fairness constraints — rather than bolted on after the fact. The review credits the rigor but tests its ceiling: technical fixes encode trade-offs someone still has to choose, and the harder questions of who sets the objective and bears the cost are governance problems no constraint can settle on its own.
Algorithmic FairnessDifferential PrivacyGovernance
Ulises A. Mejias & Nick Couldry · Nonfiction · 2024
An argument that the mass capture of human data is a new colonialism: a one-way appropriation dressed as a fair exchange, with the same logic of dispossession that justified earlier land and resource grabs. The review reads it as the missing prehistory of the AI training set — the "extraction layer" beneath model capability — and weighs its call for collective refusal against the convenience that keeps the data flowing.
Data ColonialismBig TechExtraction
Jamie Susskind · Nonfiction · 2022
A case that the central problem of the digital age is unaccountable power — over speech, attention, and the rules of online life — and that republican political theory, not just market or free-speech framing, gives the better toolkit for taming it. The review tests his proposals for codes, regulators, and democratic oversight against the harder question of who writes and enforces them once AI sets the defaults.
DemocracyPlatform GovernancePolitical Theory
Parmy Olson · Nonfiction · 2024
A dual biography of Sam Altman and Demis Hassabis that traces how two mission-driven AI labs were pulled into the orbit of Microsoft and Google, and how safety language survived contact with commercial pressure. The review reads it less as a personality story than as a governance failure: when a handful of firms set the pace, "race" logic crowds out the slow institutional work — oversight, disclosure, accountability — that durable rules require.
AI IndustryBig TechGovernance
Ethan Mollick · Nonfiction · 2024
A practitioner's field guide to working alongside generative AI, built on Mollick's "always invite AI to the table" stance and his centaur/cyborg framing. The review weighs the human-in-the-loop bargain against the "jagged frontier" his own experiments document — models that excel at some tasks and quietly fail at adjacent ones — and asks what keeping a human in the loop actually buys when the failures are hardest to see.
Generative AIHuman-in-the-LoopFuture of Work
Fei-Fei Li · Nonfiction · 2023
A memoir from one of the architects of modern computer vision, tracing the road from immigrant adolescence to ImageNet and the Stanford lab. The review reads it against the grain of the triumphant arc: the breakthrough rested on a vast, low-paid human labor of labeling and curation, and "seeing" machines inherit the choices, categories, and blind spots of the people and datasets that trained them.
Computer VisionData LaborAI History
Verity Harding · Nonfiction · 2024
An argument that the future of AI is a democratic question, not a purely technical one. Harding reads three twentieth-century cases — the space race, in vitro fertilisation and the UK's Warnock/HFEA settlement, and the governance of the early internet — for how publics, regulators, and institutions steered powerful technologies toward shared values. The review ties this to AI governance: legitimacy, public participation, and durable rules matter as much as raw capability.
AI GovernanceDemocracyHistory of Technology
Sarah Wynn-Williams · Nonfiction · 2025
An insider memoir about Facebook as a private institution with public consequences: global policy, platform governance, surveillance advertising, institutional loyalty, and the way mission language can turn avoidable decisions into scale problems. The review reads it as a prehistory of AI-agent governance, where centralized infrastructure, managed narratives, and diffused responsibility can make deployment feel inevitable before accountability catches up.
Platform GovernanceSocial MediaInstitutional Power
Emily M. Bender and Alex Hanna · Nonfiction · 2025
A sharp critique of AI hype as a language system that turns automation products into inevitability, extraction into progress, and corporate power into public destiny. The review reads it as source discipline for the age of agents: before accepting an AI claim, ask what is being automated, what evidence supports it, who benefits, who is harmed, and what recourse remains.
AI HypeBig Tech PowerGovernance
Charles Stross · Novel · 2005
A fast, dense singularity novel about AI agents, uploaded minds, corporate-personhood drift, and economies that accelerate beyond ordinary human governance. On the site it anchors cyberculture writing about runaway capital, posthuman agency, and systems that keep optimizing after human-scale meaning has been left behind.
SingularityAI AgentsPolitical Economy
Henry A. Kissinger, Eric Schmidt, and Daniel Huttenlocher · Nonfiction · 2021
An elite-policy account of artificial intelligence as a force reshaping knowledge, politics, war, institutions, and human self-understanding. The review reads it as a document of machine-mediated authority: what happens when states, firms, and universities treat AI as a new layer of cognition and geopolitical power.
AI GovernanceStatecraftMachine-Mediated Reality
Robin Hanson · Nonfiction · 2016; revised paperback 2018
A social-science forecast of a future economy built around human brain emulations: copyable minds, machine-speed work, simulated offices, subsistence competition, surveillance, identity breaks, and institutions designed around cognition as infrastructure. The review reads it as a stress test for AI labor and personhood: what happens when productive minds can be copied, priced, paused, and governed like software.
Uploaded MindsAI LaborSimulation
Tim Wu · Nonfiction · 2025
Wu extends his information-empire and attention-capture work into the economics of platform rent: the modern middleman becomes a toll collector on money, data, attention, creators, and businesses. The review reads the book as an AI-era warning that assistants, search, cloud, app stores, model APIs, and default routes can turn convenience into extraction unless interoperability, choice, evidence, and public-interest duties are built before dependence hardens.
Platform PowerAI EconomyAntimonopoly
Ray Kurzweil · Nonfiction · 1999; Penguin paperback 2000
A landmark of AI futurism about accelerating returns, machine intelligence, human-machine merger, synthetic personalities, and the possibility that future machines will be treated as conscious or spiritual. The review reads it as a belief-formation document: a place where prediction, computation, mortality, and technological hope begin to reinforce one another.
AI FuturismTranshumanismMachine Consciousness
Shannon Vallor · Nonfiction · 2024
A philosophical account of generative AI as a mirror that reflects and distorts human records, habits, values, biases, and institutions. The review reads it as a book about recursive self-understanding: what happens when people and organizations treat a machine trained on the past as a guide to judgment, agency, and the future.
AI EthicsHuman-Machine CognitionBelief Formation
Arvind Narayanan and Sayash Kapoor · Nonfiction · 2024
A field guide to separating real AI capability from inflated claims: predictive AI, generative AI, recommender systems, moderation, hype cycles, evidence, and the institutional desire to turn uncertain scores into operational truth. The review reads it as an AI-literacy book about belief formation, procurement, and the discipline of asking exactly what a system has proved.
AI HypePredictive AIInstitutional Evidence
Kai-Fu Lee · Nonfiction · 2018
An insider account of U.S.-China AI competition, deep learning deployment, data advantage, mobile platforms, entrepreneurial ecosystems, labor disruption, and the politics of implementation. The review reads it as a book about the implementation state: how AI power emerges when institutions make life machine-readable, act on the reading, and turn the changed world into new training data.
AI GeopoliticsImplementationLabor Displacement
Brian Christian · Nonfiction · 2020
A reported map of AI alignment as the gap between what machine-learning systems optimize and what humans actually meant: biased data, brittle proxies, reward design, interpretability, imitation, and value learning. The review reads it as a book about outsourced intention: how models, agents, and institutional workflows reshape judgment when objectives are compressed into data and rewards.
AI AlignmentHuman ValuesMachine Learning
Safiya Umoja Noble · Nonfiction · 2018
A foundational account of search engines as political and commercial classification systems rather than neutral windows onto knowledge. The review reads it as a warning for AI answer engines: ranking, retrieval, summarization, and generated authority can reproduce social bias while appearing merely technical.
SearchAlgorithmic BiasClassification
Yanni Alexander Loukissas · Nonfiction · 2019; paperback 2022
A data-studies book about replacing abstract "data sets" with concrete "data settings": the local instruments, institutions, interfaces, categories, and maintenance work that make data usable. The review reads it as an AI-governance manual for keeping provenance, local knowledge, and contestability attached when models turn situated records into portable authority.
Data SettingsLocal KnowledgeAI Governance
Sherry Turkle · Nonfiction · 2011; updated paperback 2017
A study of social robots, relational artifacts, phones, social media, mediated intimacy, and the loneliness that can grow inside constant connection. The review reads it as a prehistory of AI companions: machines do not need inner life to reorganize attachment, disclosure, care, and the expectations people bring back to one another.
AI CompanionsMediated IntimacyHuman-Machine Cognition
Neil Postman · Nonfiction · 1985; Penguin paperback 2005
A media-ecology classic about television, entertainment, public discourse, politics, news, education, and the cultural habits that shape what a society can treat seriously. The review reads it as an AI-era warning about feeds, chatbots, companions, and generated interfaces that can make reality continuously interesting while weakening attention, memory, and public judgment.
Media EcologyBelief FormationAI Mediation
Robert M. Geraci · Nonfiction · 2010
A religious-studies account of AI, robotics, mind uploading, virtual reality, and transhumanist salvation narratives. The review reads it as a book about the salvation loop: technical artifacts become signs of a promised world, the promised world attracts money and authority, and the new authority changes how people interpret the next artifact.
AI ReligionTranshumanismBelief Formation
Elena Esposito · Nonfiction · 2022
A sociological theory of algorithms as communication partners rather than artificial minds: machine learning, personalization, prediction, profiles, lists, and systems that produce socially meaningful outputs without human understanding. The review reads it as a grammar for AI interfaces: what happens when people, platforms, and institutions learn to communicate with systems that answer back.
Artificial CommunicationPersonalizationHuman-Machine Cognition
Meredith Broussard · Nonfiction · 2018; paperback 2019
A programmer-journalist's critique of technochauvinism: the belief that computational solutions are inherently superior. The review reads it as an AI-literacy book about institutional judgment, machine limits, legibility, automation, and the danger of treating fluent technical systems as proof that the world has been understood.
AI LimitsTechnochauvinismInstitutional Judgment
Yarden Katz · Nonfiction · 2020
A critical history of artificial intelligence as a flexible institutional ideology tied to empire, capital, university expertise, racialized models of knowledge, and carceral reform. The review reads it as an AI-era warning about the authority created when old political projects are renamed as technical futures.
AI IdeologyRacial CapitalismCarceral Systems
Kate Crawford · Nonfiction · 2021
A political and material map of artificial intelligence: minerals, labor, data extraction, classification, affect recognition, surveillance, state power, and ecological cost. The review reads it as an atlas of the machine's hidden body: the world of work, matter, institutions, and categories that vanishes behind a seamless interface.
AI PoliticsExtractionSurveillance
Tim Wu · Nonfiction · 2016
A history of attention as a market: newspapers, advertising, propaganda, broadcast media, internet platforms, and the constant attempt to get inside the human head. It belongs here because AI companions and personalized agents shift the attention economy from feed capture toward intimacy capture.
Attention EconomyPersuasionAI Companions
Michael Power · Nonfiction · 1997; paperback 1999
A foundational account of audit as a principle of social organization and control: verification rituals, auditability, internal controls, accountability demands, and the tendency of institutions to reshape themselves around what inspectors can see. The review reads it as an AI-governance warning about audits, system cards, safety cases, and compliance artifacts that can either create real contestability or become trust tokens.
AuditabilityAI AssuranceInstitutional Legibility
Virginia Eubanks · Nonfiction · 2018
An investigative account of automated welfare eligibility, homelessness triage, child-welfare risk scoring, and the digital poorhouse. It belongs here because AI governance has to ask who is first exposed to automated systems, who can appeal, and whether software expands care or merely makes scarcity easier to administer.
Welfare AutomationDigital PoorhousePredictive Risk
Aaron Benanav · Nonfiction · 2020; paperback 2022
A compact political-economic challenge to the story that robots and AI alone explain the crisis of work. The review reads it as an AI-era antidote to automation fatalism: labor markets are shaped by stagnation, ownership, institutions, bargaining power, and the stories used to make technical change feel inevitable.
AI LaborAutomation DiscourseTechnological Politics
Langdon Winner · Nonfiction · 1977; paperback 1978
A political-theory classic about the modern feeling that technical systems have escaped human control: runaway machinery, institutional dependency, complexity, technological determinism, and the stories that make political choices look like fate. The review reads it as an AI-governance book about systems that become autonomous when institutions stop treating them as choices.
Technological PoliticsRunaway SystemsAI Governance