The engineering of dependency: how the corporation that claims to serve us is reprogramming our minds. Beyond convenience: mapping the orchestrated cognitive manipulation.
When Google announces — in the polished corporate idiom particular to Big Tech — that integrating AI into search "enhances the user experience" and "simplifies access to information", it conceals beneath this rhetoric of progress an operation in psychological engineering conducted at industrial scale. We are not dealing in conspiracy theories hatched in obscure basements. We are dealing with scientifically documented mechanisms, explicitly acknowledged by former executives of these corporations who, freed from the constraints of professional loyalty, have chosen to pull back the curtain on the manipulation at work.
Tristan Harris — formerly an ethical design specialist at Google, now co-founder of the Center for Humane Technology — testified in the documentary The Social Dilemma (2020): "These platforms are intentionally designed to exploit human psychological vulnerabilities. It's not an accident. It's a strategy." Harris is no paranoid outsider. He is an insider who became a dissident.
The fundamental question is not a technological one: "Can AI generate good responses?" Obviously it can. The real question — the one the corporation systematically evades — is psychological and existential: "What does this dependency on pre-packaged answers do to our minds and to our autonomy?"
Let us deconstruct, step by step, the mechanisms by which AI integration into classical Google search does not serve us — it reprogrammes us.
The growing dependency on ready-made answers: the atrophy of cognitive muscle
The anatomy of convenience: how comfort becomes a trap
The scenario appears mundane enough: you search "flu symptoms" on Google. The classical version returned ten links to medical websites — Mayo Clinic, WebMD, peer-reviewed Romanian medical publications. You opened three or four, compared the information, evaluated the credibility of sources, formed a nuanced understanding.
Active cognitive process: evaluation, synthesis, discernment.
The AI Overview version delivers an instant, complete paragraph: "Flu symptoms include fever, muscle pain, fatigue..." Elegantly formatted bullet points. An air of absolute authority. Perfectly worded. Immaculate in presentation.
Passive cognitive process: consume, accept, close the application.
The difference seems negligible at first. Convenience, surely? Efficiency. A saving of time. But repeated fifty times a day, three hundred and sixty-five days a year, it fundamentally transforms how your brain functions.
Neuroplasticity — the brain's capacity to reshape itself according to repeated use — is no abstract concept. It is a biological reality exhaustively documented in neurology. Nicholas Carr, in The Shallows: What the Internet Is Doing to Our Brains (2010), synthesised decades of research into a single observation: "The tools we use for thinking don't merely facilitate thinking — they shape it."
When you stop practising the critical evaluation of multiple sources, that ability atrophies. Literally. The neural connections responsible for complex critical thought, left unused repeatedly, degrade. This is not a metaphor. It is neuroscience.
The corporation knows this. Google employs hundreds of specialists in behavioural neuroscience. UX designers do not guess what works — they scientifically test what maximises dependency. A/B testing with millions of users. Real-time behavioural monitoring. Constant optimisation towards maximum capture.
This is not a service. It is the engineering of addiction.
From question to command: the degradation of our capacity to articulate thought
Observe how the searches of users accustomed to integrated AI evolve over time. Initially, searches were complex and nuanced questions: "What are the benefits and risks of vaccine X for children under five with a history of allergies?"
Gradually, as the AI "learns" to anticipate and supply answers to simple queries, users devolve towards blunt commands: "vaccine children risk".
Research in cognitive linguistics demonstrates the direct correlation between the complexity of the language one uses and the complexity of one's thinking. When you stop formulating elaborate questions, you stop thinking elaborately. Language does not merely reflect thought — it constructs it.
Plato, in the Socratic dialogues, demonstrated that genuine knowledge emerges from well-formed questions, not from memorised answers. Socrates offered no answers — he compelled his interlocutors to think through successive questioning. Maieutics: the art of bringing knowledge to birth through one's own effort.
Google AI does precisely the opposite: it eliminates the necessity of effort. It infantilises the cognitive process. It transforms people who think into consumers waiting to be fed.
Saint John Chrysostom warned in the fourth and fifth centuries: "Those who always seek to be fed by others remain children their whole lives." He was speaking of spiritual dependency, but the principle applies with perfect fidelity to cognitive dependency. Intellectual maturity requires sustained effort in autonomous thought.
The erosion of source evaluation: when everything becomes equally true
False authority: the Digital Oracle syndrome
When Google AI delivers a response, it does not specify the degrees of certainty of the information it has synthesised. It does not say: "This is solid scientific consensus — based on two hundred peer-reviewed studies — but that point is contested — three studies suggest X, five suggest Y, twelve are inconclusive."
It offers The Answer. Singular. Definitive. With apparent absolute authority.
The fundamental problem: the human brain evolved for cognitive shortcuts. In ancestral tribes, when the village elder said "the red fruit is poisonous", you did not ask "What is your research methodology? Have you controlled for confounding variables?" You believed him and you survived.
Google AI exploits this evolutionary disposition. Authoritative presentation activates reflexes of automatic acceptance. Complete bypass of critical thinking.
Jaron Lanier, virtual reality pioneer and technology critic, warned in Ten Arguments for Deleting Your Social Media Accounts (2018): "Algorithms are designed to modify behaviour without the user's conscious consent. It's manipulation by design."
The logical extension: AI integrated into search is manipulation by design at the epistemological level — it modifies not only what you do, but what you believe you know, and how you come to believe it.
The collapse of the credibility hierarchy: from pyramid to soup
Before AI Overviews, users developed the ability — imperfect but functional — to evaluate credibility hierarchies:
- Level 1: Peer-reviewed academic publications, recognised medical institutions
- Level 2: Serious investigative journalism, experts with verifiable credentials
- Level 3: Specialist blogs with demonstrable track records
- Level 4: Forums, personal opinions, unauthenticated sources
A pyramidal structure. Imperfect, certainly — academics are sometimes wrong, bloggers are sometimes right. But the framework existed. The user was learning to evaluate.
AI Overview melts everything down into homogenised information soup. It synthesises equally from Mayo Clinic and from a random comment on Reddit. It processes them algorithmically, combines them, and serves the result as Unified Truth™.
Result: the collapse of the capacity to discriminate between sources of differing quality.
The adolescent who grows up this way never develops the ability to ask: "Does this information come from a credible source?" For them, everything Google says is equally true — because Google makes no visible distinction.
Jean Baudrillard, in Simulacra and Simulation (1981), described hyperreality: the stage at which representation becomes more "real" than the reality it represents. The AI synthesis becomes more "true" than the original sources — because it is more accessible, clearer, more authoritative in its presentation.
Users live inside the simulacrum. They do not know they live inside the simulacrum. They believe they are accessing reality.
Cognitive infantilisation: "I want the answer, not the process"
The instant generation: when waiting becomes intolerable
Cal Newport documents in Digital Minimalism (2019) a phenomenon he calls the progressive reduction of tolerance for sustained cognitive effort. Generations raised on instant-gratification technology develop fundamentally different neurological expectations.
A revealing comparison: ask a teenager in 2024 to read a three-thousand-word article in order to understand a complex subject. The majority abandon it after three or four hundred words. Not because they are unintelligent. Because they have not developed the cognitive musculature required for sustained effort.
Atrophied muscles do not lie — they are the predictable result of disuse. When Google AI instantly delivers the synthesis of those same three thousand words in a hundred and fifty, it eliminates the necessity of effort. The result: effort becomes progressively impossible.
The vicious circle of degradation:
- AI delivers an instant answer → the user stops practising effort
- Without practice, effort becomes harder
- Harder effort is avoided → growing dependency on AI
- Growing dependency → more pronounced atrophy
- Return to point one, amplified
A descending neurological spiral. Self-reinforcing. Potentially irreversible at the individual level — neurons lost in childhood do not miraculously regenerate in adulthood.
The illusion of understanding: when you know the answer but understand nothing
Philip K. Dick, in Do Androids Dream of Electric Sheep? (1968) — a conceptual precursor to The Matrix — explored the central problem: the difference between possessing information and understanding it.
The android in the novel can access any database instantly. It knows everything. But it does not understand in any human sense. It does not live knowledge experientially.
Adolescents raised with AI Overviews exhibit exactly this syndrome. You ask: "What caused the French Revolution?" Instant response: "Social inequality, economic crisis, the Enlightenment." Perfect. Complete. Completely uncomprehended.
They do not know why social inequality produces revolution — in some contexts it does, in others it does not. They do not understand how the Enlightenment modified mentalities — a process of centuries, not simple causality. They have the answer. They do not have the understanding.
Saint Maximus the Confessor distinguished in the seventh century between theoretical knowledge — memorised information — and experiential knowledge — understanding lived from the inside. Genuine knowledge requires personal effort in integration.
Google AI eliminates the effort. It delivers the information. It creates the illusion of understanding — the most dangerous form of ignorance, because the ignorant person who knows they do not know can still learn. The person convinced they know, whilst possessing only information without understanding, remains trapped in Plato's Cave, mistaking the shadows for reality.
The false illusion of instant comprehension: when the map becomes the territory
Confusing synthesis with comprehension
Consider the following situation — observed repeatedly in educational contexts. A student receives an assignment: "Explain Einstein's general relativity." He searches Google and AI Overview returns an elegantly formulated paragraph: "General relativity demonstrates that mass curves space-time, and this curvature produces what we perceive as gravity..."
The student copies the response. Submits the assignment. Receives a good mark. Believes he has learnt something.
Asked the following day — "Can you explain in your own words what it means for mass to curve space-time?" — silence. He cannot. He does not understand. He possesses the memorised information, not the integrated understanding.
The difference is fundamental. Genuine understanding permits:
- Transfer: applying the concept in new contexts
- Connection: linking the concept to other knowledge
- Explanation: reformulating in one's own vocabulary
- Critique: identifying limits and problems
The memorised response permits only regurgitation: reproducing the text without any cognitive process of one's own.
Google AI maximises regurgitation, minimises understanding. Not with deliberate malice — merely as the inevitable side-effect of a design optimised for maximum efficiency.
The map-mistaken-for-territory syndrome
Alfred Korzybski, founder of general semantics, enunciated the fundamental principle: "The map is not the territory." The representation of reality is not reality itself.
Users of AI Overviews systematically confuse map with territory. The AI synthesis is a map — a simplified representation, necessarily incomplete, of the complex territory represented by the original sources.
But when you never see the territory — when Google offers you only the map — you begin to believe that the map is the territory. The synthesis becomes the reality. The original sources become irrelevant.
A concrete, documented example: medical researchers observe an alarming phenomenon — patients who arrive for a consultation bearing their "Google diagnosis". They have read the AI synthesis, they have "understood" they have disease X, and they are requesting a specific treatment. The doctor explains: "Your symptoms actually suggest condition Y, which requires a completely different approach."
The patient objects: "But Google says..."
The AI synthesis becomes more authoritative than the expertise of a doctor with twenty years of experience. The map becomes more "real" than the territory the doctor is observing directly.
Baudrillard's hyperreality materialised in everyday clinical interactions.
The psychology of dependency: when you can no longer function without the crutch
Separation anxiety: the phenomenon of cognitive nomophobia
Nomophobia — the anxiety of being separated from one's phone — is already clinically recognised. The logical extension: dependency on instant AI responses generates similar anxiety when the tool is unavailable.
The empirical observation — reported repeatedly by teachers and parents: adolescents who panic when required to solve a problem without internet access. Not because they do not know how to think. Because they no longer trust their own capacity to think independently.
A psychological crutch. When you walk with a crutch for years, your legs atrophy. When you try to walk without it, you fall — not because the crutch was objectively necessary, but because prolonged dependency has atrophied the musculature.
Children raised with AI Overview will be cognitive invalids — incapable of functioning without the algorithmic prosthetic. This is not an insult. It is a prediction based on neuroplasticity and behavioural psychology.
Tristan Harris documents in The Social Dilemma: platform design is explicitly optimised for maximum dependency. This is not collateral damage. It is strategic intent. The dependent user is the profitable user.
The transfer of the locus of control: from internal to external
Clinical psychology distinguishes between:
- Internal locus of control: "Results depend on my effort and my decisions"
- External locus of control: "Results depend on external factors I cannot control"
Research consistently demonstrates: internal locus correlates with mental health, resilience, fulfilment. External locus correlates with anxiety, depression, and a sense of powerlessness.
Dependency on AI Overview progressively transfers the locus of control. From "I can find the answer through my own effort" to "The answer comes from outside; I merely wait."
Adolescents who grow up in this manner develop a foundational belief: "I cannot know anything unless Google tells me."
Abba Dorotheus, in the sixth century, warned monks against excessive spiritual dependency upon their guides — the loss of the capacity for one's own discernment through the total transfer of cognitive responsibility to an external authority. The principle is identical in cognitive dependency.
Maturity — spiritual or intellectual — requires responsible autonomy. Google AI produces the opposite: systematic infantile dependency.
The manipulation of attention: from active search to passive consumption
The attention economy: when you are the product
The internet's defining aphorism: "If the product is free, you are the product." Google does not sell search — it sells your attention to advertisers.
Alphabet's financial reports (Q3–Q4 2024) confirm: 80 per cent of revenues come from advertising. The fundamental economic model: capture attention → sell access to attention.
AI Overview is optimised to maximise time spent on the platform. When it delivers an instant answer, you do not leave for a third-party site. You remain on Google. You see more Google adverts. You generate more Google profit.
This is not theory. It is a reality documented in shareholder reports. Google executives have declared explicitly that their strategic objective is "keeping users within our ecosystem".
Translation: we will not let you leave. We keep you captive. For profit.
Shoshana Zuboff calls the system "surveillance capitalism" — the industrial extraction of human experience, converted into behavioural data, used for predictive manipulation.
They do not serve you. They exploit you. With the unconscious consent obtained through forty pages of Terms & Conditions that nobody reads.
From active agent to passive patient
The medical terminology is revealing: agent = one who acts; patient = one upon whom action is performed.
Classical search made you an agent: you decided which link to open, evaluated sources, controlled the process.
AI Overview transforms you into a patient: you receive the treatment — the answer — prepared by the authority — the algorithm — without active participation in the process.
Systematic passivisation. Nicholas Carr warned: "Tools don't merely facilitate thinking — they shape it." Google AI shapes thinking towards passivity, receptivity, dependency.
The precise opposite of genuine education, which aims at autonomy, intellectual activism, critical independence.
Conclusion: the reprogramming of minds — long-term consequences
The psychological and behavioural dimension of AI integration into search is not a footnote to the technological problem. It is the core of the problem.
The scientific documentation — from neurology (Nicholas Carr, neuroplasticity) to behavioural psychology (Tristan Harris, manipulative design) through to cognitive philosophy (Baudrillard, simulacra) — converges towards an inevitable conclusion:
Prolonged dependency on ready-made answers fundamentally alters cognitive structure.
We are not speaking of added convenience. We are speaking of neurological reconfiguration with long-term consequences:
- Atrophy of the capacity for sustained cognitive effort
- Degradation of the ability to critically evaluate sources
- Systematic confusion between information and understanding
- Transfer of the locus of control from internal to external
- Heightened anxiety when the tool is unavailable
The generation raised wholly with AI Overview will be fundamentally different cognitively from preceding generations. Not superior or inferior — differently structured.
The moral question becomes acute: do we accept this reconfiguration as inevitable progress? Or do we demand responsible separation — the preservation of classical search as an option for those who wish to maintain functional cognitive musculature?
Google has the technology for both. It has the infrastructure. It does not have the economic interest.
Separation preserves autonomous users. Forced integration produces dependent users. Dependent users generate more profit.
This is not an accident. It is an orchestration. Documented in corporate reports, acknowledged by insiders turned dissidents, exhaustively analysed by independent researchers.
The reality behind the curtain: the corporation that claims to serve you is, in fact, reprogramming you.
The final question is not "What can AI do?" It is "What do we wish to become as a species that thinks?"
For the moment, the answer to that question is being formulated by Google executives in boardrooms, optimised for quarterly profit.
We — the users and the parents of future users — can decide whether we accept or contest that reality.
Primary documentary sources:
- Tristan Harris — The Social Dilemma (2020), Center for Humane Technology
- Nicholas Carr — The Shallows: What the Internet Is Doing to Our Brains (2010)
- Cal Newport — Digital Minimalism (2019)
- Jaron Lanier — Ten Arguments for Deleting Your Social Media Accounts (2018)
- Shoshana Zuboff — The Age of Surveillance Capitalism (2019)
- Jean Baudrillard — Simulacra and Simulation (1981)
- Philip K. Dick — Do Androids Dream of Electric Sheep? (1968)
- Alphabet Inc. — Financial reports Q3–Q4 2024
- Cognitive neuroplasticity research — multiple peer-reviewed studies
- Saint Maximus the Confessor, Abba Dorotheus — patristic writings on knowledge and autonomy
Essential Questions
from "The Psychological and Behavioural Dimension — the Prohibition of Critical Thinking"
1. What concrete difference exists between classical Google search and AI Overview from a cognitive standpoint?
Classical search returned ten links which the user opened selectively, compared sources, and formed an understanding through their own effort — an active cognitive process. AI Overview delivers an instant synthetic paragraph, definitive and sourceless — a passive cognitive process. Repeated fifty times a day, the former trains the brain; the latter atrophies it.
2. What is neuroplasticity and why does it matter in this context?
Neuroplasticity is the brain's capacity to structurally reconfigure itself according to repeated use — neural connections that are called upon strengthen; those that are neglected degrade. Nicholas Carr demonstrates in The Shallows (2010) that the tools of thought do not merely facilitate thinking — they shape it. If you cease practising critical evaluation, that ability atrophies literally, not metaphorically.
3. What did Tristan Harris say, and why does his voice carry more weight than that of an external critic?
Harris was an ethical design specialist inside Google itself — not an academic commentator. He states explicitly in The Social Dilemma (2020): "These platforms are intentionally designed to exploit human psychological vulnerabilities. It's not an accident. It's a strategy." The weight of this testimony lies in his position as insider, not in critical theory.
4. How does AI degrade the capacity to formulate questions?
Users accustomed to immediate responses devolve from complex enquiries — "What are the risks of vaccine X for children with allergies?" — to telegraphic commands: "vaccine children risk". Research in cognitive linguistics reveals the direct correlation: the complexity of the language one uses both reflects and constructs the complexity of one's thinking. When you cease formulating elaborately, you cease thinking elaborately.
5. What is maieutics, and why does Google AI do precisely the opposite?
Maieutics is the Socratic method of "delivering" knowledge through successive questions that compel the interlocutor to think independently — knowledge emerges through one's own effort, not through memorised responses. Google AI eliminates the effort: it supplies the finished answer, infantilising the cognitive process. It transforms people who think into consumers waiting to be fed.
6. What is the Digital Oracle syndrome?
When AI delivers a response without specifying degrees of certainty — without distinguishing between solid scientific consensus and active controversy — it activates evolutionary reflexes of automatic acceptance. The human brain is evolutionarily configured for cognitive shortcuts: when the "village elder" speaks, you do not request a methodology. Google AI exploits this reflex. The result: complete bypass of critical thinking.
7. How did the credibility hierarchy function before AI Overview, and what has happened to it?
There was a functional pyramid: peer-reviewed publications at the summit, serious journalism below, specialist blogs, unauthenticated opinions at the base. Imperfect, but structured — the user was learning to evaluate. AI Overview melts everything into "homogenised information soup", synthesising equally from Mayo Clinic and from an anonymous comment on Reddit, serving the result as Unified Truth™. The collapse of discrimination between sources.
8. What does Baudrillard describe through "simulacra", and how does it apply to AI?
In Simulacra and Simulation (1981), Baudrillard analysed the stage at which representation becomes more "real" than the reality it represents — hyperreality. The AI synthesis becomes more "true" than the original sources because it is more accessible, clearer, more authoritative. Users live inside the simulacrum without knowing they live inside it — they mistake the map for the territory.
9. What is the descending neurological spiral and why is it dangerous over the long term?
AI delivers an instant answer → the user stops practising effort → effort becomes harder → effort is avoided → growing dependency on AI → more pronounced atrophy → return to point one, amplified. This self-reinforcing spiral may become irreversible at the individual level: neurons lost in childhood do not regenerate miraculously in adulthood.
10. What distinction does Saint Maximus the Confessor draw between theoretical and experiential knowledge?
In the seventh century, Saint Maximus distinguished between knowledge as memorised information and knowledge that is lived experientially, integrated through personal effort. Genuine knowledge requires active participation, not passive reception. Google AI eliminates precisely this effort — it delivers information without understanding, creating the most dangerous form of ignorance: the ignorant person convinced they already know.
11. What is the map-mistaken-for-territory syndrome and how does it manifest clinically?
Korzybski formulated the principle: "The map is not the territory." The AI synthesis is a map — a simplified representation of the complex reality held in the original sources. Patients who arrive at their doctor's with a "Google diagnosis" and contest clinical expertise are the concrete expression of this confusion: the algorithmic map becomes more authoritative than the observation of a physician with twenty years of practice.
12. What is cognitive nomophobia?
Classical nomophobia — the anxiety of being separated from one's phone — is already clinically recognised. The logical extension described in this essay: dependency on instant AI responses generates similar anxiety when the tool is unavailable. Teachers and parents report adolescents who panic when faced with problems without internet access — not because they cannot think, but because they no longer trust that they can think independently.
13. How does design-for-dependency function in platform architecture?
Corporations employ specialists in behavioural neuroscience who do not guess what works — they scientifically test what maximises dependency through A/B testing with millions of users and constant optimisation. The dependent user is the profitable user. Design is explicitly optimised for maximum capture, not for the user's benefit.
14. What distinguishes internal from external locus of control, and why does it matter?
Clinical psychology distinguishes between an internal locus — "results depend on my effort" — which correlates with mental health, resilience, and fulfilment, and an external locus — "results come from outside me" — which correlates with anxiety, depression, and a sense of powerlessness. Dependency on AI Overview progressively transfers the locus from internal to external: "I cannot know anything unless Google tells me." Intellectual maturity becomes impossible.
15. What warning does Abba Dorotheus give about spiritual dependency, and how does it apply cognitively?
Abba Dorotheus, in the sixth century, warned monks about the loss of the capacity for their own discernment through the total transfer of cognitive responsibility to an external authority — the spiritual director. The principle is identical in algorithmic dependency: maturity — spiritual or intellectual — requires responsible autonomy. Google AI produces the opposite: systematic infantile dependency.
16. What does it mean that the user has moved from "agent" to "patient"?
The medical terminology is revealing: an agent acts; a patient is acted upon. Classical search made you an agent — you decided what to open, you evaluated, you controlled. AI Overview transforms you into a patient — you receive the algorithmic treatment without participating in the process. Systematic passivisation. The precise opposite of what genuine education aims to achieve.
17. What is "surveillance capitalism" in Shoshana Zuboff's definition?
Zuboff defines the system as the industrial extraction of human experience, converted into behavioural data and used for predictive manipulation. Google does not sell search — it sells your attention. Alphabet's financial reports confirm: over 80 per cent of revenues come from advertising. AI Overview is optimised to keep you inside Google's ecosystem, not to serve your intellectual needs.
18. What long-term consequences does this essay foresee for the generation raised with AI Overview?
The atrophy of the capacity for sustained cognitive effort; the degradation of critical source evaluation; the systematic confusion between information and understanding; the transfer of the locus of control from internal to external; heightened anxiety when the tool is unavailable. This generation will be fundamentally different cognitively — not inferior or superior, but structurally reconfigured for dependency.
19. Why does Google not offer the option of separating classical search from AI?
Google has the technology and the infrastructure for both. Separation would preserve autonomous users. Forced integration produces dependent users. Dependent users generate more profit. This is not negligence — it is a strategic decision, documented in corporate reports and acknowledged by insiders turned dissidents.
20. What moral question does the author formulate in the conclusion?
The question is not technological — "what can AI do?" — but existential and moral: "What do we wish to become as a species that thinks?" For the moment, the answer to that question is being given by Google executives in boardrooms, optimised for quarterly profit. Users and parents of future users can decide whether they accept or contest that reality.
Author's Note: limitations, tensions, comparisons, and rhetorical tools
Acknowledged limitations of the argument
This text constructs a powerful argument, but it has vulnerabilities which I recognise and own. The first: neuroplasticity works in both directions — the brain atrophies through disuse, but it also restructures itself under new stimulation. There is no scientific consensus that the cognitive damage produced by digital dependency is irreversible on the dramatic scale suggested here; formulations such as "potentially irreversible" or "cognitive invalids" are the rhetoric of warning, not validated clinical diagnoses.
The second vulnerability: the text treats users as a homogeneous mass. There are adolescents who use AI precisely as a point of departure for deeper research — not as a substitute for it. The tool does not determine behaviour; educational context, family, and the habit of reading determine how the tool is used. The argument applies to the unreflective average user, not to the complete universe of users.
The third: the sources selected — Harris, Carr, Newport, Lanier, Zuboff — form a coherent chorus, but not a neutral one. All are critics of digital technology; the voice of the optimists — who exist and have arguments — is deliberately absent from the construction. This is advocacy rhetoric, not a systematic review of the literature.
Internal tensions
There is an unresolved tension between the patristic argument and the neuroscientific one. Saint Maximus the Confessor and Abba Dorotheus operated within a thoroughly spiritual anthropology, in which cognitive autonomy and discernment are conditions of salvation — not autonomous humanist values. Neuroplasticity, on the other hand, is a naturalistic mechanism without teleology. The essay combines them as though they were saying the same thing. They say something similar — the necessity of personal effort — but from ontologically incompatible frameworks. The tension is not resolved; it is masked through juxtaposition.
Another tension: the text accuses Google of infantilising the user, whilst simultaneously employing a rhetoric that leads the reader step by step, without allowing them to draw their own conclusions. "This is not a service. It is the engineering of addiction." "They do not serve you. They exploit you." These phrases are definitively formulated, without any remainder of doubt. There is a paradox here — a text defending critical thinking written in a manner that discourages it.
Comparisons and counterpoints
The comparison with Plato's Cave is perhaps the most fertile in the text — and the most under-exploited. Plato described prisoners who mistake shadows for reality and who, once freed, return to persuade the others and are killed for it. The parallel with the AI user's situation is rich: who liberates them? Who is Socrates in the algorithmic age? The essay does not push this comparison to its conclusion.
Philip K. Dick is invoked briefly. Do Androids Dream of Electric Sheep? is not "the conceptual basis for The Matrix" — it is a distinct novel with its own concerns about empathy and humanity. Reducing it to a precursor of The Matrix is a simplification that does a disservice both to Dick and to the reader.
Baudrillard is probably the best-used reference — AI-synthesised hyperreality corresponds faithfully to the mechanism of his simulacra. The example of the patient who disputes the doctor with "But Google says..." is the most convincing moment in the text, precisely because it is not abstract: it is concrete, documented, and everyday.
Rhetorical tools employed
The text works with several recurring instruments. The first is bipolar contrast: active versus passive cognitive process, agent versus patient, map versus territory, knowledge versus information. Binary oppositions are pedagogically effective, but they reduce complexity — the cognitive-digital reality is far more gradual than harsh contrast allows.
The second instrument is the accumulation of authority: Harris + Carr + Newport + Lanier + Zuboff + Baudrillard + Plato + Maximus the Confessor + Abba Dorotheus. The accumulation creates the impression of unanimous convergence. The risk: the diversity of sources may conceal the genuine theoretical differences between them.
The third is the rhetoric of exposure: "The corporation knows this." "This is not theory. It is documented reality." "It is not an accident. It is an orchestration." The repetition of these formulas creates the effect of curtains being pulled back, of hidden truth brought to light. It is an effective rhetorical gesture — and, as noted above, slightly paradoxical in a text that defends precisely the suspicious thinking that questions authorities claiming to possess the truth.
The fourth — probably the most subtle — is the approach to the reader through "you": "your brain", "it reprogrammes you", "we will not let you leave". Direct address in the second person singular transforms the abstract argument into a personal charge. The reader is implicated, not a spectator. It is a technique of accountability — but also of rhetorical pressure.
This remains one of the most coherent and best-documented arguments I have constructed on this subject. It is not perfect, and it makes no claim to neutrality. It is advocacy that is conscious of itself — and precisely for that reason, it requires this note.