Through McLuhan’s Lens

Regulation Outpaces Skills Vocabulary

September 07, 2026 | 2820 words

Through McLuhan’s Lens: The Word We Reach For, and the Words We’ve Lost

When a Society Learns One Verb

Listen to any conversation about artificial intelligence right now, and you will hear the same word surface again and again. Regulate. We debate how to regulate AI, when to regulate it, whether we are regulating it fast enough or too fast. The word has become the reflex, the automatic hand-motion of a nervous public. It sounds responsible. It sounds like the adults are in the room.

But something is missing from the room, and this week’s data lets us name it. When you tally the language society actually uses to talk about governing AI, one word towers over the rest while its natural companions barely register. We say “regulate” fluently and constantly. We say “collaborate,” “partner,” “build skills with,” and “transform alongside” haltingly, if at all. The vocabulary is lopsided. And a lopsided vocabulary is never neutral. It has already decided something for us, quietly, before any citizen cast a vote.

This essay is about that quiet decision. Not the content of any particular rule, but the shape of the conversation itself. Because the conversation is doing something to us that no one intended, and almost no one has noticed.

The Numbers Behind the Reflex

Start with the imbalance itself, because it is stark. The discourse around AI governance is saturated with the language of control and starved of the language of participation. Regulation-talk dominates. Skills-talk, partnership-talk, and transformation-talk trail far behind, appearing as afterthoughts when they appear at all. We are assembling an elaborate rulebook for a technology we have not yet learned to hold in our hands.

Think about how strange that is. Imagine a civilization drafting detailed laws about automobiles before anyone had learned to drive one. The laws would be full of confidence about speed limits and liability. They would be silent about steering, because no one had felt the wheel. That is roughly where we stand. The pieces analyzing this moment keep circling the same gap. The Great AI Governance Debate documents a public square where the machinery of rulemaking runs at full volume while the vocabulary of working with the technology stays nearly mute. The imbalance is not a rounding error. It is the defining feature of how we talk.

And here is the load-bearing point, the thing the numbers actually prove. This is not a temporary lag, a case of the partnership-language simply running a few years behind. A discourse that can say “regulate” easily but “collaborate” only with effort has already chosen a relationship. It has cast humans as the governors and AI as the governed, the public as the ruled-over rather than the hands-on. The vocabulary gap is not a problem to be caught up on later. It is shaping the relationship right now, in advance, while we assume we are merely describing it.

To see why, we need a better instrument than common sense. We need to look at the conversation not as a window but as a thing in itself.

The Medium Is the Message

Marshall McLuhan built his career on a single unsettling idea, and it is the key that opens this whole problem. He called it “the medium is the message.” In plain English: the tool we use to communicate shapes us more than whatever we happen to say through it. The content is the shiny distraction. The real action is in the form itself — what the medium does to our habits, our senses, our sense of who we are.

McLuhan pressed this point hard in Understanding Media. A light bulb, he noted, carries no “content” at all, yet it utterly reorganizes the spaces and hours of human life. It lets us hold meetings at midnight and play ball games after dark. The message of the light bulb is not any particular thing it illuminates. The message is the new pattern of living it creates. What matters is not what flows through the medium but what the medium does to the shape of our days.

Now apply that lens to AI governance, and the whole scene tilts. We keep arguing about content. Which rule? How strict? Applied to which companies? These feel like the important questions, and they consume nearly all the oxygen. But McLuhan’s idea reframes the case entirely. The message of our governance discourse is not any specific regulation. The message is regulation itself as a posture — the stance toward AI that the word installs in us every time we reach for it.

And what is that stance? Regulation is the language of distance. You regulate a thing you stand apart from, a thing you manage from a control room, a thing you fence rather than befriend. When “regulate” becomes the only civic verb available for AI, the word quietly teaches us our position. We are the authorities at the console, keeping watch over something out there. We are not the drivers. We are not the collaborators. We are the wardens.

The medium of our conversation — its dominant vocabulary — is delivering that message beneath every debate about content. You can win every argument about which rule is best and still lose the thing that matters, because the very act of framing AI as a thing-to-be-regulated has already settled what kind of creatures we will be in relation to it. Managers at a distance. Never hands on the wheel.

Who Benefits From the Console

Before we go further, a reader-serving question. When a single word crowds out all its rivals in a public conversation, it is worth asking who gains from the arrangement. Language does not tilt on its own. Somebody is usually more comfortable when it tilts a certain way.

The regulation-frame serves a specific set of interests, and none of them is you. It serves regulators, whose relevance depends on there being something to regulate and whose authority grows with every new rulebook. It serves the growing industry of compliance vendors — the consultants and software firms who sell you the tools to prove you followed the rules. Where there is regulation, there is a market in navigating it, and that market has every reason to keep the conversation locked in the language it profits from.

And it serves large institutions managing risk. For a corporation or an agency, “regulate” is a comfortable word. It converts a bewildering new technology into a familiar category: a compliance problem, a box to check, a liability to be contained. The regulation-frame lets powerful institutions treat AI the way they treat everything — as something to be processed by existing machinery.

Notice who is missing from that list. The ordinary citizen who will live under AI systems every day. The public gains nothing from being cast as the governed. When the only verb on offer is “regulate,” the citizen is handed a purely spectator role. You watch the authorities manage the thing. You do not touch it, shape it, or learn to work it yourself. The vocabulary keeps you in the audience while telling you that your safety is being handled. That is a comfortable message to receive and a costly one to accept.

The Rear-View Mirror

Why did we reach for “regulate” in the first place? Why does that word feel so natural while “collaborate with AI” feels awkward on the tongue? McLuhan offers a second concept that answers the question precisely, and it is one of his sharpest.

He called it the “rear-view mirror.” The idea is simple and slightly humiliating. When something genuinely new arrives, we cannot see it directly. So we name it and manage it using the categories of the world that just ended. We march backward into the future, eyes fixed on the mirror, describing the road ahead in the vocabulary of the road behind. In The Medium Is the Massage, McLuhan put it bluntly: we see the present through a rear-view mirror, and we march backwards into the future. We attach the new to the familiar because the familiar is all our language has ready.

The history of technology is a graveyard of rear-view names. The automobile was the “horseless carriage.” Radio was “wireless.” We named the new thing by the old thing it was replacing, because the old thing was the only shape our minds could grasp. The name always lagged the reality, and the lag was not harmless — it shaped how we treated the new arrival for years.

“Regulate” is a rear-view mirror word, and this is the heart of the matter. It comes from the age of factories and broadcasts. You regulate a smokestack. You regulate a power plant, a railroad, a television signal. These are the technologies of the industrial and broadcast eras — big, centralized, standing at a distance from ordinary people, pumping out a fixed product that citizens merely receive. Regulation was the perfect civic verb for that world. The public could not operate a power grid, so the public’s role was to demand it be regulated. Fair enough.

But AI is not a smokestack. It behaves nothing like a broadcast tower. It is not a centralized thing pumping out a fixed product at you from a distance. It is a technology you can converse with, argue with, direct, and shape through your own participation. It responds to how you hold it. It is closer to a musical instrument or a collaborator than to a power plant. And yet we have reached, reflexively, for the smokestack word. We are naming the horse-and-buggy age’s horseless carriage all over again, describing an interactive, participatory technology in the vocabulary of the passive broadcast era.

McLuhan’s warning about the rear-view mirror was never that old words are quaint. It was that old words are blinding. They let us feel we understand a thing while ensuring we do not. Every time we say “regulate AI,” the smokestack ghost inside the word tells us AI is something to stand back from and fence off. The word smuggles in a picture of the technology that is simply wrong — and the wrong picture then governs how we relate to the real thing.

The Hot Frame That Leaves No Room

There is a further edge to this, and McLuhan gives us the tool for it too. In Understanding Media, he divided media into “hot” and “cool.” A hot medium is high-definition. It fills in every blank and hands you a finished, saturated product, leaving nothing for you to do but receive it. A cool medium is low-definition. It leaves gaps you must fill yourself, and so it demands your participation. Hot media make you a spectator. Cool media make you a collaborator.

The regulation-frame is scorchingly hot. It is a high-definition vocabulary that fills in every blank for you. The rules are drafted, the categories are set, the enforcement is specified. Your part is to receive the finished product and be reassured. Nothing is left open. Nothing invites your hands.

The language of skills and partnership is cool by contrast. To talk about building skills with AI, or partnering with it, or transforming your work alongside it, is to open a space that only your participation can fill. These words assume you will be a doer, not a spectator. They leave gaps for your involvement precisely because involvement is the point.

This is why the vocabulary imbalance matters at the level of the senses, not just the dictionary. A discourse dominated by hot regulation-language actively suppresses participation. It offers a finished world and asks you to sit still inside it. The starvation of the cool, participatory words is not an accident of tone. It is the natural consequence of a hot medium doing what hot media do — filling every blank so completely that there is no room left for the public to lean in.

The Revelation: A Society That Governs But Cannot Hold

Now we can name the thing the discourse is doing that no one intended.

Put the pieces together. The medium is the message: our governance vocabulary installs a posture of distance regardless of any particular rule. The rear-view mirror: we borrowed “regulate” from the smokestack age, and it blinds us to what AI actually is. The hot frame: regulation-language fills every blank and leaves no room for us to participate.

The result is a civilization training itself into a very specific and very strange condition. We are becoming fluent governors of a tool we never learned to hold. We are perfecting the language of management-at-a-distance for a technology that only reveals itself to those who get close and put their hands on it. We are building an elaborate console for a machine we have refused to touch.

This is the invisible decision the vocabulary made for us. Not “how should AI be governed” — that question is at least out in the open, argued about, contested. The buried decision is this: the human relationship to AI will be one of distant control rather than hands-on partnership. No one debated that. No one voted on it. It arrived smuggled inside a single overused word, delivered beneath the content of every argument, while we thought we were merely discussing which rules to write.

McLuhan’s most useful move — the one that serves the reader more than any other — was what he called reversing figure and ground. The figure is the thing everyone stares at, the object in the spotlight. The ground is the environment around it, the context nobody notices precisely because it is everywhere. In the AI debate, the figure is the endless argument over specific rules. The ground is the vocabulary itself — the fact that “regulate” is the water we swim in. Everyone stares at the figure. The revelation is in the ground. The word doing the most to shape our future is the one we have stopped hearing.

And a society that can govern but cannot collaborate has a name for its condition. It is estrangement. You end up ruling something you never came to understand, managing from a control room a technology you might have learned to hold as a partner. You get all the anxiety of responsibility and none of the fluency of skill. You become the warden of a country you have never walked through.

What Citizens Lose, and What They Could Demand

So what is actually at stake for you, the reader who will live under AI systems rather than design them?

When “regulate” is the only civic verb available, you lose your role as a participant. You are handed a seat in the audience and told the management is being handled by experts at the console. Your safety, you are assured, is their department. But safety-by-management is a thin thing. It keeps you protected in the way a fence keeps cattle protected — from a distance, without your knowledge, and without your consent to anything but the fence itself. You never learn the technology. You never gain the fluency that would let you shape it. You remain the governed.

A fuller vocabulary would give you back your hands. If “skills” sat alongside “regulation” in the public conversation, you could demand not just that AI be fenced but that citizens be equipped to use it — that the public gain the fluency now reserved for a technical priesthood. If “partnership” carried equal weight, you could ask what a genuine collaboration between humans and these systems looks like, rather than accepting a relationship of distant supervision as the only option on the menu. If “transformation” were spoken as fluently as “regulation,” you could insist on a say in how your work and your community are reshaped, instead of merely being told the reshaping will be safely managed.

These are not softer demands than regulation. They are stronger ones. They refuse the spectator’s chair. They insist that the public be a hand on the wheel and not a passenger reassured that the driver is licensed. Regulation asks only to be protected. Partnership asks to be involved. The vocabulary that can only say the first has already surrendered the second.

McLuhan’s whole project, from Understanding Media onward, was to make people aware of their environments before those environments finished shaping them. He believed that once you see the ground you have been ignoring, you regain some power over it. The medium stops being your master the moment you notice it is a medium at all. That is the reader-serving heart of his work, and it is the whole point of this column.

So here is the practical inheritance of this week’s data. The next time you hear “we must regulate AI,” you will hear something you could not hear before. You will hear the smokestack ghost inside the word. You will hear the distance it installs, the spectator’s chair it hands you,