Through McLuhan’s Lens

The Expertise Inversion

May 31, 2026 | 2679 words

Through McLuhan’s Lens: The Hollowing of the Expert

There is a moment now repeating itself in classrooms, offices, and clinics across the country, and it goes something like this. A senior person — a professor, a manager, a physician — asks a junior person to “just run it through the AI.” The junior person does. The output comes back polished, structured, fast. And then a small, telling thing happens: the senior person cannot fully evaluate what they are looking at. They can judge the content, perhaps. But they cannot judge the process that produced it, because they do not have it in their hands the way the junior person does. The authority to assign the task and the fluency to perform it have come apart. The person above no longer fully understands the work of the person below.

This is the Expertise Inversion, and the surface debate around it is loud and familiar. Are students cheating? Are faculty falling behind? Should we ban the tools or mandate them? These are the questions the discourse keeps asking. They are not wrong, exactly. They are simply aimed at the wrong layer. They stare at the figure — who is more skilled — and miss the ground, which is what the medium has already done to the meaning of skill itself.

Marshall McLuhan is dead, but the instrument he built for seeing this kind of thing is not. Read through Understanding Media, the inversion stops being a story about clever kids and lagging adults. It becomes a story about a medium quietly rewriting the social category of expertise while everyone argues about who has more of it.

The Message Is Not the Essay

Start with the phrase McLuhan is most known for and most misunderstood by: the medium is the message. It means that the form of a technology reshapes us more powerfully than any particular content the technology carries. The railway did not introduce movement into human society, but it changed the scale and pace and pattern of everything it touched. The content riding on the rails — coal, passengers, mail — mattered far less than the new shape of life the rails imposed.

Apply that to AI fluency. The entire surface debate is a debate about content. Did the student write this paragraph or did the machine? Is this output original or derivative? Is it true or hallucinated? These are content questions. They assume the important thing is what comes out of the tool.

But the medium is the message. The thing reshaping the social order is not any individual essay or memo or diagnosis. It is the form of AI fluency — the new pattern of who can summon expert-grade output on demand, and how fast, and at what scale. That pattern is rearranging authority regardless of what any single prompt produces.

Consider what the data shows about where the fluency actually lives. Surveys this year found that a majority of students report regular use of generative AI for academic work, while only a minority of faculty report comparable fluency with the same tools. One widely cited figure put student adoption at well over half, with faculty trailing far behind — in some samples, fewer than one in five instructors described themselves as confident users. Hold those two numbers next to each other. They do not describe a skills gap. They describe an inversion of the medium’s center of gravity. The people with the least institutional authority have the most fluency with the dominant medium. The people with the most authority have the least.

The content debate cannot see this, because the content debate is busy asking whether the students’ essays are honest. The medium-is-the-message lens asks a different question: what happens to a hierarchy when the medium of competence migrates downward and the medium of authority stays put?

The answer is that the hierarchy starts running on a fiction. The expert still occupies the expert’s chair. But the expertise — understood now as fluency with the operative medium — has left the chair and walked across the room.

What AI Fluency Extends, and What It Numbs

McLuhan’s second great instrument is the idea of media as extensions of man. Every technology, in Understanding Media, extends some human faculty outward — and in the same gesture, numbs or amputates another. The wheel extends the foot and numbs the leg’s labor. The book extends the eye and numbs the ear. Extension and amputation are not separate events. They are the same event seen from two sides.

So ask the McLuhan question directly. What does AI fluency extend in the student, and what does it numb in the expert?

In the fluent novice, AI extends reach. It extends the capacity to produce structured, expert-sounding work far beyond the novice’s own internalized knowledge. A second-year student can now generate a literature review, a legal-sounding brief, a passable diagnosis-shaped paragraph. The faculty of synthesis — once the slow earned reward of years of training — is extended outward into the tool. The novice reaches past their own competence.

This is genuinely powerful, and the column will not pretend otherwise. But McLuhan’s law is symmetrical. What extends must also numb. And here the data offers a warning the vendors selling “AI fluency” never print on the box. Studies this year measuring cognitive engagement found significant drops in critical scrutiny among heavy AI users — a tendency to accept fluent output without the friction of independent verification. The extension of reach comes bundled with a numbing of the very judgment that would tell you when the reach has overshot the truth. The student gains the output and loses the part of themselves that would have checked it.

Now turn the lens on the expert, because this is where the inversion does its quietest damage. The expert is numbed too, but differently. The expert’s amputation is not of judgment but of centrality. For a generation, the expert’s authority rested on a simple monopoly: they held knowledge the novice could not quickly get. That monopoly was the medium of their authority. AI fluency dissolves the monopoly. The novice can now summon, in seconds, output that approximates what the expert spent decades acquiring.

McLuhan called the response to this kind of over-extension technological numbness — the self-amputation that follows when a faculty is overwhelmed. Understanding Media describes how we go numb precisely in the area a new medium most affects, because the shock is too great to feel directly. The expert, watching their monopoly evaporate, does not experience it as a clear loss. They experience it as irritation, as suspicion, as the urge to police. The bans, the detection software, the moral panic about cheating — these are not strategies. They are the numb reflex of an amputated authority, flailing at the figure because it cannot feel the ground.

This is why the discourse runs so hot on the question of cheating and so cold on the question of what expertise now is. The cheating frame lets the numbed expert keep the old map. It says: the hierarchy is fine, the categories are intact, we merely have rule-breakers to catch. It is a frame that protects the chair by refusing to notice the expertise has left it.

The Rear-View Mirror, and the Hierarchy We Keep Steering By

McLuhan had a precise image for this refusal. He called it the rear-view mirror — the human habit of seeing the present through the frame of the immediate past. “We look at the present through a rear-view mirror,” he said in the McLuhan Interview. “We march backwards into the future.” We attach the new to the old, name the car a “horseless carriage,” and steer by an image that is already behind us.

The expertise hierarchy is now a rear-view image. Expert above, novice below; the one who knows certifies the one who is learning. That structure was built for a world where knowledge moved slowly and authority pooled in the people who had accumulated the most of it. It was a sensible map for that world. It is a rear-view mirror for this one.

Watch how the rear-view mirror operates in the live debate. Both sides — the institutions defending the hierarchy and the technologists celebrating its disruption — are looking at the same mirror. The institutions see students who must be governed back into the old order. The technologists see students who have “leapfrogged” their teachers. Both descriptions assume the old axis is still the real one: that there is a clear above and below, and the only question is who occupies which slot. They are arguing about seating in a room whose architecture has already changed.

The number that exposes the mirror is the adoption gap itself — the majority of students fluent, the minority of faculty confident. Read through the rear-view lens, that gap looks like a temporary lag, a problem to be closed by training the laggards. But the medium is the message. The gap is not a lag in a stable hierarchy. It is the visible edge of a hierarchy turning inside out. Steering by the mirror, the institutions keep treating an inversion as a delay. They keep prescribing faculty development for what is actually a redefinition of expertise as a social medium.

And here is the figure-ground reversal McLuhan’s whole method exists to perform. Pull the background forward. The thing everyone is staring at — who is more skilled — is the figure. The ground, the thing that makes the figure possible and that no one is looking at, is this: the medium has already redefined what skill means.

What the Discourse Cannot See About Itself

This is the revelation the column has been building toward, and it is metacognitive — it is about what the argument itself is doing.

When the institution says “students are cheating” and the technologist says “students are ahead,” they appear to disagree completely. But notice what they agree on. Both assume expertise is a possession — a thing you either have or lack, hold in your head or fake with a machine. The whole quarrel is about the rightful distribution of this possession. Who really has it. Who only pretends to.

The medium has hollowed out that assumption beneath both of them. In a world of AI fluency, expertise is no longer best understood as a stored possession. It is becoming a relationship — a capacity to direct, interrogate, and verify a medium that holds the stored knowledge for you. Expertise is migrating from knowing the answer to knowing what to ask and how to test what comes back. That is not the old expertise diminished. It is a different faculty entirely, running on a different medium.

This reframes everything. The student who scores high on adoption but low on critical scrutiny is not “more expert” than the cautious professor. And the professor who knows the field but cannot operate the medium is not securely “more expert” than the student. Both descriptions use a word — expert — that the medium has quietly emptied. The two sides are fighting over a category the technology already redefined while they argued. They are dividing the contents of a box that no longer holds what the label says.

McLuhan understood that media work this way precisely because they are invisible to their users. In The Medium Is the Massage, the recurring point is that environments created by new media are not noticed as environments. They are felt as the natural condition of things. The fish, in the old line, knows nothing of water. The participants in the expertise debate are swimming in a redefinition of expertise they cannot see, because it is the water. They experience it only as a content problem — cheating, accuracy, fairness — never as the medium problem it actually is.

This is also why the loudest voices benefit from keeping the figure in view and the ground hidden. The vendor selling “AI fluency training” needs you to believe expertise is still a possession you can purchase and stack. The institution defending its credentials needs you to believe the hierarchy is real and merely under attack. Both have an interest in the rear-view mirror. The reader does not. The reader is served only by pulling the ground into view and naming it plainly: the medium changed the meaning of the word everyone is fighting over, and the fight is a way of not noticing.

The Citizen Inside the Inverted Hierarchy

This is not, finally, a campus story. The classroom is just where the inversion became visible first, because schools are where society sorts the expert from the novice most formally. But the same inversion is arriving everywhere authority rests on a knowledge monopoly — which is to say, nearly everywhere.

The junior associate who can produce the brief faster than the partner can check it. The new hire who out-operates the manager on the tool that now runs the department. The patient who arrives having interrogated the same medical literature the doctor relies on. In each case the old vertical — expert above, novice below — is being crossed by a new horizontal: fluency with the medium, distributed without regard for rank. McLuhan saw this re-tribalizing pattern coming. He argued that electronic media collapse hierarchy into network, dissolving the centralized, top-down order of the print age into something closer to a global village — knowledge held in connection rather than in command.

For the ordinary citizen, this is the part that matters, and it cuts two ways.

The hopeful side is real. When fluency with a powerful medium spreads downward and outward, gatekeeping weakens. Capacities once locked behind credentials and years of training become available to people who never had access to them. That is a genuine democratization, and no defense of the old hierarchy should be allowed to pretend otherwise. The numbed expert’s reflex to police is not a neutral act. It is often a monopoly protecting itself.

But McLuhan’s law is symmetrical, and the citizen should hold both sides at once. The same studies that show fluency spreading show critical scrutiny thinning. A society where everyone can summon expert-grade output, and fewer and fewer people retain the friction-born judgment to test it, is not a society that has abolished the need for expertise. It is a society that has hidden the need while raising the stakes. The danger is not that machines will replace experts. The danger is that fluency will outrun judgment everywhere at once — that we will become collectively better at producing authoritative-sounding answers and collectively worse at knowing which ones to trust.

That is what it means when fluency with a medium outruns the authority structures built to govern it. The structures do not adapt gracefully. They go numb, then defensive, then irrelevant — while the medium reorganizes the social order underneath them. We will live for some time inside hierarchies that are formally intact and functionally inverted: experts in the chairs, fluency in other hands, and a public unsure whose authority to trust because the word expert no longer points where it used to.

The reader’s defense is not a tool or a policy. It is the figure-ground move itself — the refusal to mistake the loud surface fight for the quiet structural change beneath it. When the next round of the debate arrives, and it will, the question to carry is not who is more skilled. That question accepts the hollowed category. The question is: what is the medium doing to skill itself, and who profits from keeping that invisible?

McLuhan is gone. But the instrument holds. It was built to drag the ground into the light while everyone stares at the figure — and a public that learns to use it will not be managed by the inversion. It will see it coming, name it plainly, and decide for itself what expertise should mean in a world where the machine knows the answer and the human, increasingly, must remember how to ask.