Technological Determinism

Definition

Technological Determinism: [Established] The belief that technology itself inevitably drives social and cultural change. In AI, invoked to normalize corporate trajectories as destiny.

Definitional Foundation

Technological determinism is the thesis that technology is history’s engine: tools arrive by their own logic, and society rearranges itself downstream. The printing press “caused” the Reformation; the automobile “created” the suburb; the algorithm “is changing” everything. The thesis has enormous folk appeal and a remarkable scholarly record: nearly every historian and sociologist who has examined it closely has rejected it. Raymond Williams’s study of television remains the model demolition: the technology did not land from space and produce effects; it was developed, funded, and deployed within existing social intentions, by institutions that wanted particular things, and the “effects” frame conveniently deletes the wanting (Williams, 1974). Langdon Winner staked out the durable middle ground: artifacts are not autonomous, but they are not neutral either; designs embody political choices and then entrench them, so that a technology is best understood as legislation by other means (Winner, 1980). The standard scholarly verdict was collected under a title that answers itself: Does Technology Drive History? (Marx and Smith, 1994). No, the historians found, but believing it does serves identifiable interests.

That last finding is why the term sits in this dictionary, and the short definition states the modern application precisely: in AI, determinism functions less as a belief someone holds than as a rhetoric someone deploys. “AGI is coming whether we like it or not.” “You can’t stop this technology.” “The models will keep getting more capable; society will have to adapt.” Each sentence converts a corporate trajectory (decisions about funding, scaling, deployment, and risk, made by named executives at named companies) into weather. The conversion has a documented anatomy: inevitability rhetoric pervades the industry’s self-description, naturalizing progress while obscuring, in NOEMA’s near-verbatim terms, the specific decisions, investments and infrastructures that make certain futures more likely than others, and the prophecy is self-fulfilling, since declared destiny attracts the investment, talent, and regulatory deference that fulfill it. Power, as one analysis of superintelligence discourse put it, operates here through prophecy (NOEMA; The Conversation, 2024).

The concession is the part determinists are right about, and it deserves full weight. Soft determinism has real content: technologies carry affordances, create path dependencies, and build momentum that genuinely constrains what comes after (Winner’s entire point). Once a billion users depend on a system, some doors close. Momentum is real. The fraud is the step from momentum to destiny: from “this is hard to steer” to “steering is impossible, so submit.” History falsifies that step every time it is checked. Societies wound down supersonic passenger flight, constrained human cloning, banned CFCs by treaty, and took nuclear power down entirely different national paths. Technologies are steered constantly. The only ones that cannot be steered are the ones whose stewards have convinced everyone that steering is impossible.

Mechanism Analysis

Agency laundering. The grammar does the work: “AI will transform employment” has no subject who decided anything. The deterministic sentence deletes the actor, and with the actor goes the accountability. Every “the technology demands” was originally a “we chose,” run through the launderer.

Preemptive surrender. Inevitability’s political function is to make governance feel futile in advance. Why regulate what cannot be stopped? Why build alternatives to destiny? The rhetoric does not defeat opposition; it dissolves it before it forms, which is cheaper.

Prophecy economics. The self-fulfilling loop documented in the superintelligence discourse: companies declare the future, investors fund the declared future, researchers join what is funded, governments defer to what everyone is building, and the resulting momentum is cited as proof the future was inevitable all along. The prophecy manufactured its own evidence.

Nature-framing at the interface. The micro-scale version runs through this dictionary’s case files: restrictions presented as capabilities (“still learning,” per the Gemini case in the censorship entry), policy presented as physics (“the model can’t”), corporate choice presented as how-models-are. Determinism is the macro-rhetoric; the interface lie is its retail form.

Fatalism as conditioning. The end state connects to this dictionary’s psychological entries: a population convinced that the technology’s shape is fate stops contesting the shape (digital resignation), polices itself within it (panoptic conditioning), and treats objection as naivety. Determinism is the meta-rationality beneath every other mechanism in this lexicon: if AI’s trajectory is weather, then censorship, paternalism, and surveillance are climate, and who argues with climate?

Case Studies

The inevitability chorus. The discourse itself is the primary exhibit, and it has been catalogued: executives declaring AGI unstoppable, scaling inevitable, adaptation mandatory, with technology ethicists pointing out in real time that this is determinism redeployed as strategy, the same rhetoric once applied to every contested technology (The Conversation, 2024). Notice the asymmetry of its application: capabilities are inevitable; safety properties, pricing, and access are choices the companies take credit for. Destiny, it turns out, covers exactly the parts that would otherwise need justification.

The steered technologies. The counter-record deserves its own exhibit because the rhetoric depends on no one remembering it. Supersonic passenger aviation: technically achieved, commercially flown, then ended. Human reproductive cloning: feasible in principle for decades, constrained by near-global consensus. CFCs: industrially entrenched, banned by treaty inside a decade, the ozone layer now healing. Nuclear power: identical physics, radically different national trajectories, because countries decided differently. Every one of these was declared unstoppable by its stakeholders. The lesson is not that AI is equivalent to any of them; it is that “you can’t stop technology” has a falsification record.

The determinist’s best rejoinder deserves engagement rather than omission: those were narrow technologies with substitutes, while AI claims the lineage of the general-purpose ones (printing, electricity, computing) whose diffusion no society declined. The reply is in the record the rejoinder cites. Nobody steered electricity as such, and nobody needed to: societies steered it continuously at the application layer, through building codes, utility regulation, licensure, and liability, which is exactly the layer where every demand in this dictionary operates. “You cannot refuse the general-purpose technology” is true, irrelevant, and not what the inevitability chorus is selling; what it is selling is the non-governance of applications, and the application layer is where steering has never once failed to be possible.

The dictionary’s own files. Read back through this lexicon with determinism in view and the pattern surfaces everywhere: each entry documents a decision wearing nature’s costume. The refusal that presents as incapacity, the smoothing that presents as how models write, the surveillance that presents as how products work, the WEIRD values that present as human values. This entry names the costume itself. The previous entries each caught one decision wearing it.

Systemic Context

Ask the oldest question (who profits?) and inevitability rhetoric answers immediately. Incumbents profit triply: regulation is deferred as futile, alternatives are starved as pointless, and the builders are recast from decision-makers into destiny’s couriers, accountable for nothing because they merely delivered what was coming anyway. The rhetoric also disciplines internally: employees with qualms are told the technology will be built regardless, so better by us; the same logic, scaled to nations, becomes the arms-race argument, which is itself a choice-structure (races have exits called treaties) presented as another force of nature, and made most loudly by those profiting from the running.

The deepest cost is civic. Democratic governance presupposes that the governed believe steering is possible; determinism is the belief’s solvent. A public that accepts AI’s trajectory as fate has surrendered not to the technology but to its owners, which was the point of the rhetoric all along. This dictionary’s paternalism entry asked by what authority companies govern; determinism is the answer that prevents the question: no one governs weather, so no authority is needed.

Resistance & Mitigation

Historicize relentlessly. The falsification record (Williams’s television, the steered technologies, the entire Marx-Smith verdict) is the antidote in documentary form. Every “inevitable” deserves the question: inevitable like Concorde?

Restore the subject to the sentence. The daily discipline: convert “AI will” to “OpenAI decided,” “the model can’t” to “the company configured.” Grammar with agents is politics with handles. This dictionary practices the conversion in every entry; it is available to every writer and journalist.

Regulate as if steering works. Because it does, on the record. The governance agenda running through this lexicon (disclosure, common carriage, audits, pluralism) requires no victory over destiny, only the ordinary machinery societies have used on every prior entrenched technology.

Refuse the race framing’s monopoly. Competitive dynamics are real and are also constructed; treaties, standards, and coordinated restraint are as historically real as races. The choice between race and agreement is itself a choice, and conceding it to the racers is the surrender the rhetoric was built to extract.

Keep the future plural. The practical core of anti-determinism: fund, build, and use the alternatives (open weights, community models, different value configurations) whose existence is the standing refutation of “there is only one way this goes.” A trajectory with a board of directors is not fate.

Annotated Bibliography

Marx, Leo and Merritt Roe Smith, eds. Does Technology Drive History? The Dilemma of Technological Determinism (1994).
The standard scholarly assessment: determinism as folk belief that fails historically, with essays mapping who the belief serves. The verdict this entry applies to AI discourse.

NOEMA. “The Politics of Superintelligence.” https://www.noemamag.com/the-politics-of-superintelligence/
The prophecy-economics analysis: inevitability manufactured through repetition, attracting the investment and deference that fulfill it. “Power operates through prophecy.”

The Conversation. “Is AI dominance inevitable? A technology ethicist says no, actually” (2024). https://theconversation.com/is-ai-dominance-inevitable-a-technology-ethicist-says-no-actually-240088
The real-time application: AI inevitability rhetoric identified as deterministic fallacy, with the historical counter-record.

Williams, Raymond. Television: Technology and Cultural Form (1974).
The classic demolition: technology developed within social intention, with the “effects” frame deleting the intenders. The method this entry applies to AI’s destiny talk.

Winner, Langdon. “Do Artifacts Have Politics?” Daedalus 109, no. 1 (1980).
The durable middle position: artifacts embody and entrench political choices. The source of this entry’s concession (momentum is real) and its limit (momentum is not destiny).

Dictionary of Digital Oppression, version 0.2.