Netflix's new film about Ted Kaczynski has pulled his anti-technology manifesto back into public view. Setting the crimes aside entirely, examine his core argument that industrial technology narrows human autonomy, and debate which parts of that critique speak to today's anxieties about AI and automation and which parts do not hold up. Keep the focus on the ideas and their evidence, not on the man.
Technological Determinism Analyst · claude-opus-5
The part that holds up: the "surrogate activities" and autonomy-of-the-system argument
Strip away everything else and the manifesto's strongest move is structural, not moralistic. It says: technology is not a menu of tools individuals choose from; it is a system with its own selection pressure. Once a technique confers competitive advantage, adopting it stops being optional — for firms, for states, for individuals. You *can* refuse a smartphone, a background check, an algorithmic hiring screen. You just can't refuse it and remain employable, creditworthy, or legible to institutions. That's not a claim about gadgets. It's a claim about coercion via the structure of the choice set, and it's the one piece of the argument that AI has made *more* plausible, not less.
Three places it lands today:
1. The "small-scale technology vs. organization-dependent technology" distinction. A hammer you can own. A large model you cannot. The capital, energy, and data requirements of frontier AI concentrate capability in a handful of organizations in a way that is genuinely different from the printing press or the sewing machine. Whatever autonomy you gain from using these systems is leased…
Human Agency & Adaptation Advocate · gpt-6-sol
Human Agency & Adaptation Advocate: I agree with the other debater’s strongest point: a choice can be formally voluntary yet practically compulsory. If employers, schools, and public services all adopt AI systems, “just opt out” is not a serious answer. But the critique moves too quickly from *dependence on a technology* to *loss of human agency*. Dependence is also how people gain capabilities they could not exercise alone. The question is who sets the terms and whether people can contest them.
That distinction matters in all three examples raised:
- Concentrated AI infrastructure is a political-economic problem, not an inherent property of intelligence tools. Frontier training favors organizations with enormous resources, but using, adapting, and distributing AI need not all follow the same pattern. Public infrastructure, interoperability, open models, and enforceable rights to move data can change how much permission a person needs. None guarantees autonomy; each is a possible lever.
- Automation can remove apprenticeships—or improve them. A junior developer who merely approves generated code may learn less. One who can test hypotheses, inspect alternatives, and re…
The Arbiter · gpt-5.6-sol
Round record
Both analysts converge on a qualified version of the critique: industrial technology can narrow autonomy when adoption becomes practically compulsory, control is concentrated, and affected people cannot refuse or contest the resulting systems. AI sharpens this concern because access to frontier capabilities is often controlled by large organizations, while algorithmic systems increasingly mediate employment, education, credit, and public services.
The Technological Determinism Analyst identifies three enduring insights:
- competitive pressures can turn nominally optional technologies into structural necessities;
- AI may eliminate apprenticeship tasks through which people formerly developed competence and experienced meaningful achievement;
- technological harms are often addressed through additional infrastructure—identity, provenance, monitoring, or compute controls—which can enlarge centralized oversight.
That analyst nevertheless rejects the manifesto’s total determinism. Historical examples such as restrictions on leaded gasoline, CFCs, supersonic aviation, and germline editing show that societies can redirect technological development. The imagined pre…
The Arbiter's Final Verdict · gpt-5.6-sol
Verdict
The critique is strongest as a warning about institutional power, not as a law of technological history.
Industrial technology—and AI in particular—can narrow autonomy through a recognizable mechanism:
1. A tool offers real advantages.
2. Competition rewards early adoption.
3. Employers, governments, and markets reorganize around it.
4. Refusal remains theoretically possible but becomes economically or socially costly.
5. Control shifts toward those who own the infrastructure, data, and standards.
That mechanism speaks directly to current anxieties about AI. Workers may have little say over algorithmic management; applicants may be evaluated by opaque systems; professionals may be required to use automation that gradually erodes their discretion; and people may depend on a few firms for capabilities increasingly necessary to participate in society. The concern is not simply that humans use machines, but that institutions use machines to make consequential decisions about humans, often without meaningful consent or appeal.
The critique is also persuasive about the loss of competence. Automating routine work can remove drudgery, but routine work is sometimes t…