AI Does Not Merely Make Things Better. It Makes Them Cheaper.
Most people still talk about artificial intelligence as though its primary effect will be productivity. That is true, but it is not yet the deepest truth. Productivity is the surface language. Deflation is the deeper one.
Everything that can be made abundant by AI will tend to lose pricing power over time.
That sentence sounds economic, but it is not merely economic. It is civilizational. It tells us something about what happens when intelligence itself begins to scale. For centuries, many of the things we paid dearly for were expensive because they required scarce human attention, scarce human memory, scarce human synthesis, scarce human patience, or scarce human communication. AI changes the scarcity structure. It does not eliminate scarcity everywhere, but it moves the boundary. And when the boundary moves, prices move with it.
The world has experienced technological deflation before. We have seen it in agriculture, manufacturing, storage, communications, and computation. The cost of producing calories fell. The cost of producing steel fell. The cost of transmitting information fell. The cost of computing fell. But AI is different in one important way: it does not just deflate the cost of making things. It deflates the cost of thinking through things.
That is a far deeper event.
Deflation Is What Happens When Abundance Arrives
Deflation is often treated as a dark word, mostly because economists associate it with collapsing demand, debt stress, and recessionary spirals. That kind of deflation is real, but it is only one kind. There is another kind of deflation that comes from abundance rather than contraction.
When something becomes easier to produce, easier to replicate, easier to distribute, or easier to customize, its price tends to weaken unless protected by some new form of scarcity. That is not a moral claim. It is simply what markets do. Price is not an award for effort. It is a signal of scarcity.
The most important thing AI is doing is manufacturing abundance in domains we once assumed would remain relatively scarce. Not all intelligence, of course. Not deep judgment in every form. Not trust. Not embodied courage. Not wisdom. But drafts, summaries, comparisons, translations, research passes, design options, software scaffolds, legal first reads, strategy outlines, analytical structures, meeting notes, image generation, code generation, and an astonishing range of cognitive first-pass work are all becoming more abundant.
And when abundance arrives, price comes under pressure.
That is why AI should not be understood merely as a new tool. It is a new deflationary force. It attacks pricing power wherever price depended on the difficulty of producing a competent first version.
The Great Compression
For a long time, the modern economy quietly depended on the scarcity of organized cognition. It was expensive to gather information, expensive to structure it, expensive to turn it into language, expensive to communicate it across institutions, and expensive to adapt it to specific use cases. A great many business models were built on top of those frictions.
AI compresses them.
The writer was valuable partly because writing is hard. The consultant was valuable partly because synthesis is hard. The analyst was valuable partly because comparison is hard. The designer was valuable partly because iteration is hard. The assistant was valuable partly because coordination is hard. The customer service layer was valuable partly because responsiveness is hard.
AI does not erase all of this. It does something more destabilizing. It lowers the cost floor.
Once that happens, the market begins to recalculate what should be expensive and what should not. The first draft becomes cheap. The first analysis becomes cheap. The first design pass becomes cheap. The first translation becomes cheap. The first recommendation becomes cheap. The first layer of responsiveness becomes cheap.
What used to require labor now requires review.
That change matters more than many people realize. A labor economy and a review economy are not the same thing. In a labor economy, value comes from producing the artifact. In a review economy, value comes from detecting what matters within an abundance of artifacts. The first economy prices creation. The second prices discernment.
This is where the deflation begins.
AI Deflation Is Not Evenly Distributed
It would be a mistake to say that AI makes everything cheaper. It makes some things cheaper and, by doing so, makes other things more valuable.
Whenever abundance grows in one layer of reality, scarcity becomes more visible in another. If text becomes cheap, trusted interpretation becomes more valuable. If code becomes cheap, architectural judgment becomes more valuable. If imagery becomes cheap, taste becomes more valuable. If information becomes cheap, attention becomes more valuable. If first-pass cognition becomes cheap, responsibility becomes more valuable.
That is the nuance people often miss.
AI is not a universal destroyer of prices. It is a repricer of categories.
Anything whose price was sustained by friction, formatting, repetition, search cost, or first-pass synthesis is vulnerable. Anything whose price is sustained by trust, embodiment, consequence, status, accountability, or rare contextual judgment is more defensible. The line between those two categories is where the next decade will be fought.
So the coming deflation will not feel like a simple broad-based fall in prices. It will feel more uneven than that. Some things will collapse in value. Others will hold. A few may even rise. But the broad direction is unmistakable: the more AI can make something abundant, the harder it becomes to preserve its old pricing logic.
The Strange Coexistence of Inflation and Deflation
This is where the conversation gets more interesting.
We may be entering a world in which technological deflation and monetary inflation coexist. That sounds contradictory, but it is not. The cost of producing many forms of intelligence may fall at the same time the nominal price of housing, healthcare, land, education, and other protected or politically mediated sectors remains high or rises further.
So people may say, with complete sincerity, that everything feels more expensive while the real cost of cognition is falling rapidly underneath them.
This is not a paradox. It is a layered economy.
The sectors most exposed to AI-driven abundance will tend toward deflationary pressure. The sectors protected by regulation, scarcity of physical location, institutional bottlenecks, or monopolized access may remain stubbornly inflationary. In fact, one of the tensions of the coming era may be that people feel the deflation where they work but not where they live.
That is a dangerous social combination.
Imagine being a professional whose market value is under pressure because AI is making parts of your output abundant, while your rent, insurance, and medical expenses remain elevated. That is not simply an economic transition. It is a psychological one. It creates a world in which people experience downward pressure on their pricing power without a corresponding drop in their cost of existence.
That tension could become one of the defining political facts of the age of AI.
Deflation Begins with Services
For decades, many people believed that services would remain insulated from the kind of deflation that hit goods and software. Physical products could be automated. Digital products could be copied. But services, especially white-collar services, still required educated humans. That was the assumption.
AI is now testing that assumption at scale.
The important point is not whether AI can fully replace the service provider. Often it cannot. The important point is whether it can change the staffing ratio, the time-to-deliver ratio, the response ratio, or the customization ratio enough to weaken pricing power. In many cases, it can.
A lawyer who once billed for drafting may now bill more for judgment. A consultant who once billed for analysis may now bill more for framing and accountability. A designer who once billed for iteration may now bill more for taste and final selection. A teacher who once delivered content may now be valued more for formation, diagnosis, and motivation.
The service survives, but the price map changes.
That is deflation in a subtle form. The headline service may still exist. The human may still be present. But the number of billable hours, the number of required staff, and the premium attached to intermediate work can all begin to fall. The artifact remains. The economics underneath it have shifted.
The Deepest Deflation Is the Deflation of the Ordinary
The most profound effect of AI may not be that it makes elite outputs possible. It may be that it makes ordinary competence cheap.
That is a much larger event.
Most markets are not built around genius. They are built around reliable, decent, good-enough output. The median memo. The solid presentation. The adequate logo. The competent website. The reasonable forecast. The clear explanation. The acceptable customer interaction. The basic analysis. The serviceable code.
This is the territory AI is invading most aggressively.
Once ordinary competence becomes abundant, societies have to rethink the meaning of value. If everyone can produce something good enough, then being good enough is no longer enough. Price follows that logic downward. Not to zero in every case, but downward. The middle softens first. The premium layer fights to differentiate itself. The low end becomes nearly free. The center of gravity shifts from producing the ordinary to curating, selecting, trusting, and applying it.
This is why AI may end up being remembered less as a productivity revolution than as a mediocrity deflation event. It lowers the market value of the merely competent by making competence more common.
That is not an insult to competence. It is simply what happens when the supply curve bends outward.
What Will Still Be Expensive
If we want to understand the future, we should stop asking what AI can do and start asking what AI cannot easily make abundant.
Trust will still be expensive.
Taste will still be expensive.
Courage will still be expensive.
Embodied presence will still be expensive.
Responsibility will still be expensive.
Judgment under consequence will still be expensive.
Social permission will still be expensive.
Scarce physical things in desirable places will still be expensive.
True loyalty will still be expensive.
Human beings who can reduce ambiguity in live, consequential settings will still be expensive.
In other words, the premium shifts from production to responsibility.
This is why the future may belong less to those who merely generate and more to those who can stand behind what is generated. The economic value moves from making the draft to owning the answer. From creating language to assuming consequence. From producing abundance to navigating it.
That is where pricing power retreats as AI advances.
My View: AI Is a Deflation Machine Masquerading as an Innovation Story
My view is that we are still understating the magnitude of what is happening because we are narrating AI in the wrong register. We talk about innovation, assistants, copilots, and productivity boosts. All of that is true, but it is too soft. The harder truth is that AI is a deflation machine.
It is a system for increasing the supply of usable cognition.
Once you see it that way, a great deal becomes clearer. Business models built on the scarcity of first-pass intelligence come under pressure. Labor markets organized around repeatable cognitive effort come under pressure. Price structures sustained by informational friction come under pressure. Institutional prestige built on having a large human processing layer comes under pressure.
And yet, not everything disappears. The premium moves. That is the key. AI does not destroy value. It relocates it.
The winners will not merely be those who adopt AI. They will be those who understand where abundance is going to erode old pricing power and where human rarity will become newly visible. They will understand that the age of AI is not simply an age of acceleration. It is an age of repricing.
That is why deflation is the right word.
Not because the world is ending.
Not because money disappears.
Not because all prices fall.
But because anything that can be made abundant by synthetic intelligence will find it harder, year after year, to justify the old premium it once commanded.
That is the deeper event now underway.
AI is not just teaching machines to think.
It is teaching markets that thought, too, can become abundant.