Simulacra of Thought: How AI Created a New Aristocracy of Questions and Killed the Process of Understanding (Or What to Teach Children Today)

Simulacra of Thought: How AI Created a New Aristocracy — Those Who Know How to Ask Questions

Prologue: The Era of Flawless Simulacra and the Death of Process

We build cathedrals out of emptiness. Around us — majestic facades of strategies, arches of business plans, stained-glass windows of reports. Each document is an architectural masterpiece, flawless in its structure, harmonious in composition, precise in wording. Modern masters — algorithms — have learned to imitate the gothic of thought with such perfection that distinguishing the handcrafted from the generated is already impossible.

But step inside these cathedrals. Instead of living ideas — bare walls. Instead of dialogue — echo. Instead of thinking — perfect form, concealing a fundamental absence of content. These documents smell of seriousness, but they are sterile. They were not born in the throes of search — they appeared ready-made, like Athena from the head of Zeus, bypassing the stages of pregnancy, labor, the pain of birth.

This is a cultural catastrophe that has already occurred. We have taught machines to imitate the results of intellectual work, completely destroying its process. And it is precisely in the process — in that very “digging” with a consultant, on marker boards stained with diagrams, in nightly arguments and sudden insights — that true understanding was born. Understanding not as a set of theses, but as a reassembled worldview, as a new optic for looking at a problem.

Today, a manager comes to a meeting with a ready-made cathedral in hand. “Our 140-page strategy,” he says, and confidence is read in his eyes. Confidence not that he has understood the problem, but that he has a heavy, substantial file. He did not walk the path — he teleported to the final point. And it seems to him that he has conquered the summit, though in reality he was simply delivered there by cable car, bypassing all the difficulties of ascent.

We have created a civilization of flawless simulacra. AI-bureaucracy is the apotheosis of this civilization. If before, access to real discussion of a problem was blocked by people with their own interests and prejudices, now it is blocked by perfectly packaged voids — documents that create an impenetrable illusion that thinking is no longer necessary. The problem is “solved” by beautiful text. A checkmark can be placed, and one can move on. The question “where to?” drowns in the roar of self-satisfaction from fulfilling a formal requirement.

Act I: Historical Amnesia — How We Fear the Wrong Thing Again

Every technological prophecy of intellectual degradation repeats the same scenario. When in the 15th century Gutenberg invented the printing press, Europe’s educated elite wrung their hands in horror: “People will forget how to remember! Knowledge will become cheap and accessible to any ignoramus! Oral culture will die!” From the height of five centuries, we see: printing did not destroy memory — it freed it for more complex tasks. It did not kill culture — it created mass culture and the scientific revolution.

We are at the same point again, only the technology is different. “AI will make people stupid!” cry modern Cassandras. But let’s look at the mechanics of this fear through the prism of cognitive economics.

Let’s take another, less obvious example — cartography. Once, the ability to navigate by stars, read terrain, and memorize routes was a vital skill. The emergence of accurate maps, and then GPS, would seem, should have led to the degradation of spatial thinking. What actually happened?

Before: 90% of mental effort — on navigation (“where am I? where to go?”). 10% — on comprehending space (“what’s happening here? why is it arranged this way?”).

After: 2% — on entering an address into a navigator. 98% — on analyzing the urban environment, planning logistics chains, designing routes considering multiple parameters (time, cost, safety, aesthetics).

GPS did not simplify navigation — it made it more complex. It raised the bar from “how to get there” to “how to build an optimal system of movements.” Exactly as the calculator once raised the bar from arithmetic to mathematical modeling.

AI goes through exactly the same evolution, but with thinking as a whole. It takes not intellect, but first-order intellectual routine:
– Mechanical writing of texts by template
– Manual collection and primary aggregation of data
– Composing plans from ready-made conceptual blocks

The real danger is not even the loss of these skills. The danger is that, having gained the ability to instantly generate “solutions,” we will forget how to recognize real problems. We risk getting a generation of managers capable of ordering a brilliant strategic plan from AI, but unable to answer a simple question: “What client pain does this plan actually address?”

The machine answers questions phenomenally. But it is fundamentally incapable of doubting the question. Incapable of saying: “Wait, are we asking the right question? Maybe the problem is elsewhere?” And it is precisely in this ability that the essence of human reason lies. The difference between AI and a human is like between an ideal taxi driver who will quickly take you where you say, and a wise guide who can stop you and ask: “Are you sure you want to go there? Maybe we should take a different path?”

Act II: The Hierarchy of Cognitive Competencies — What Remains When Knowledge Became Infrastructure

Subject knowledge has died. It hasn’t disappeared — it has become a utility service. Like electricity or plumbing. A doctor who memorizes the symptoms of all diseases is today — an anachronism. A marketer who knows all traffic channels — an archaic. Knowledge has descended to the level of an operating system. It must work flawlessly, but it goes unnoticed.

So what rises to the top? A new map of competencies:

Level 1 (Historical): Execution Skill.
What it was: Ability to write, calculate, make plans.
What AI does: Takes it completely. Does it faster, cheaper, often better.
Status: Became hygiene. Like the ability to read — necessary but insufficient.

Level 2 (Transitional): Subject Knowledge.
What it was: Deep understanding of one’s field.
What AI does: Makes it instantly accessible. Democratizes expertise.
Status: Became infrastructure. Like knowledge of the alphabet — you can’t do without it, but you can’t live by it alone.

Level 3 (New Elite): Engineering of Thinking.
What it is now:
1. Ability to Problematize — turning vague unease into a clearly formulated problem. Not “we need to increase sales,” but “what fundamental dissatisfaction of our client are we ignoring?”
2. Architecture of Context — designing systems in which AI tools work meaningfully. Determining: which data is relevant, how to collect it, how to verify the result.
3. Selection of Meaning — distinguishing deep insight from statistically plausible nonsense. AI generates all possible versions. The person chooses the single one that has meaning.

A person becomes not an executor, but a task-setter and supreme judge. Your job is not to write a strategy, but to:
– Create a process in which the right questions are born
– Configure a conveyor where AI processes data, and people evaluate the conclusions
– Make the final decision based on what cannot be algorithmized: wisdom, intuition, values, ethical considerations

You are the director. AI is the ideal, but Intent-lacking operator, lighting technician, and editor. It will do everything technically flawlessly. But only you know what film you want to shoot.

Act III: Education in the Era of Omniscient Machines — What to Teach When Homework Is Done by a Neural Network?

Traditional pedagogy is built on a simple scheme: knowledge transmission → skill training → verification of assimilation. Homework is the cornerstone of this system. What happens when any student gets a perfect solution from AI in 30 seconds?

We find ourselves at a crossroads:
– The Path of Prohibition (fighting AI, controlling, blocking) — this is a path to nowhere. The technology cannot be banned. This is fighting the tide.
– The Path of Capitulation (allow everything, let AI learn for the children) — this is a path to catastrophe. We will get a generation skilled at formulating requests but incapable of thinking.

There is a third way — the path of reassembling education around a new goal.

What to teach in 2026?

1. Discipline of Thinking as Mental Hygiene.
– Meta-cognitive Practices: Teach children to observe the course of their own thoughts. Not “did I solve it correctly?” but “how did I arrive at this solution? where could errors in logic have crept in?”
– Philosophy as a Tool: Not the history of ideas, but the practice of questioning. The ability to break down any statement into components: what premises? what assumptions? what follows?
– Working with Uncertainty: Most real tasks have no correct answer. Need to teach how to act under conditions where the criterion of correctness is unknown.

2. The Art of Problematization — Seeing the Task Behind the Assignment.
Instead of “solve the equation 2x+5=15” give: “We have 15 conditional units of resource. Each operation of type A consumes 2 units, system setup — 5 units. How many operations A can we perform? What if the resource can be increased? What if each operation gives a different result?”

The student must:
– Extract the essence from the context
– Formulate several different versions of “what is even needed here?”
– Choose criteria for evaluating the solution
– And only then use AI to test hypotheses

3. Skills of Dialogue with AI — Not as with a Search Engine, but as with a Colleague.
This is not about “how to write a prompt,” but about meaningful interaction:
– Iterative Deepening: The first answer is the beginning of a conversation. “What if we look from another angle?”, “Under what conditions will this stop working?”
– Context Management: “Explain as if I were an investor/client/competitor”, “Consider constraints: time X, budget Y, legal framework Z.”
– Critical Verification: “Where does AI know this from?”, “What data could have misled it?”, “How to verify this conclusion independently?”

4. Value Orientation — Ethics as a Practical Skill.
AI has no values. It optimizes a given function. The key skill is to be aware of one’s values and anticipate the consequences of their application. Ethics ceases to be a subject, becoming a tool for designing decisions.

The Exam of the Future — Not a Test, but a Defense of the Process:
1. The student chooses a real problem from their surroundings.
2. Designs research: how to study it, what data is needed, how to involve AI.
3. Conducts analysis, critically evaluating the results.
4. Proposes a solution, explicitly arguing for it from the standpoint of chosen values.

Act IV: Symbiosis Instead of Competition — The New Ecosystem of Thinking

The implementation of AI is not a technological challenge, but an existential opportunity. We are forced to answer anew: what is our uniqueness as a species?

The answer lies in the division of cognitive labor:

AI — the cognitive digestive system.
Processes enormous volumes of information “food.” Filters out noise, identifies patterns, finds correlations. Quickly, efficiently, without fatigue.

Human — the cognitive nervous system.
Sets goals. Experiences doubts. Makes choices based on meanings that are not reducible to data. Feels responsibility. Seeks beauty in a solution.

Practical Steps for the New Reality:

1. Revision of KPIs in Business.
Stop measuring value by the volume of documents produced. Introduce metrics:
– Quality of questions asked
– Depth of reflection on errors
– Ability to identify “blind spots” in data

2. Creation of “Meaning Factories.”
Instead of analytics departments — problematization laboratories. Their task is not to give answers, but to find new questions. To doubt the obvious. To seek contradictions in successful processes.

3. Institutionalization of Doubt.
Implement into the decision-making process a mandatory stage: “request to AI for counterarguments.” Ask the neural network to find weak points in any plan, identify risks not considered.

4. Development of a Culture of Questioning.
The most scarce resource of the future is not data, but the ability to ask transformative questions. This should become a key competence of a leader, teacher, parent.

Epilogue: Return to the Only Important Work

AI has done an amazing thing. It did not take work away from us — it returned to us the most important work. The one from which we ran for centuries into the hustle of execution.

The machine freed us from the humiliating role of good executors. It returned to us the heavy burden of being thinkers.

That very cat after the vet… Perhaps it lost something secondary to gain something primary — the opportunity to live in a new role. Not a wild hunter, but a companion. Not a reproductive machine, but a personality.

So our flawless, empty PDFs are a symptom of a painful but necessary transition. We are burying the illusion that we think when we simply structure the known. We are forced to begin thinking for real — at the level of formulating problems, choosing values, making decisions under conditions of fundamental uncertainty.

AI will not make us stupid. It will make us, finally, adults. It will force us to admit: our only competitive advantage is the ability to ask questions for which machines have no and cannot have answers. The ability to feel the pain of a problem even before it is formulated. The willingness to sit before a blank sheet, experiencing reverent horror at ignorance, and — despite this horror — begin to dig. Together. Until we get to the essence.

Discipline of thinking — that is the new currency. And it cannot be generated. It can only be earned — slowly, painfully, without guarantees. This is precisely what we will have to do in the coming decades. Return to the only work that has always been ours: to think. Not instead of the machine. Not in spite of it. But as only a human can — holistically, responsibly, doubting and wondering.

© Tatyana Burmagina & EWA