Age of Economics
The World Does Not Have Markets
Much of what we learn in economics is misleading because the world does not have markets. Real markets barely exist outside of financial exchanges, where standardized products trade on centralized platforms with real-time pricing, many participants, negligible search costs, transparency and market forces.
What we actually have is a system designed to maximize what we pay without us realizing it.
Take Gucci. Gucci is not in the handbag market. Gucci does not want to be in the handbag market. The handbag market is a commodity market, and no one at Gucci has any interest in competing on commodity terms. Gucci is in the Gucci handbag market — a market of one, for one product, where they set prices at whatever level they want.
Gucci is the dream for every company: one supplier, high demand, no competition.
Peter Thiel understood this when he declared that competition is for losers. In a world of perfect competition, profits go to zero. Every successful business is therefore essentially a monopoly — it has found something unique to do, something that escapes the sameness of competition. The goal is not to win the competition. The goal is to make the competition irrelevant.
But Thiel was describing the aspiration. He was less interested in how most companies actually pull it off. The primary mechanism for winning nowadays is not building something great. It’s winning over your psychology.
The Machine
McDonald’s arch is visible from the highway. The small restaurant next door can’t afford a sign. That’s not an accident. That’s the system working.
Logos on clothes turn people into billboards. Packaging is designed so your eyes land before your brain does. Search results are flooded with ads, so the links you think you chose were chosen for you. Influencers don’t recommend things — they manufacture need. Landing pages convert before you can think. If you’re a company, the game is simple: be the only brand anyone sees.
None of this is a market. It’s a machine built to prevent market forces from ever kicking in.
And killing real markets is so profitable that it corrupts everything it touches. Google alters its own search results to make advertised products look competitive — even though those products have to cost more, because someone has to pay for the ads. The news is garbage because clicks beat nuance, and nuance doesn’t sell ads. Even the comparison sites and blogs people trust to help them shop are taking kickbacks on the sales they drive. The entire apparatus of modern commerce is designed to make sure you never encounter a real price signal.
The Age of Markets
But AI doesn’t care about your psychology. It knows more than any human, and it can be told to work for the buyer — not the seller.
AI is the beginning of the end of the age of psychology, and the start of the age of markets.
When the fog clears, every product, every company, every industry gets sorted into two categories: those whose distinctiveness is real, and those whose distinctiveness was always a fiction held together by exploiting how people think. Companies will have to compete by building better products. The news will have to sell news instead of ads. Consumers will have power and clarity again.
But AI won’t commoditize everything, and it won’t kill advertising. Not for lack of ambition — for fundamental reasons.
What AI Cannot Touch
Some products are subjective in ways that resist comparison entirely. A dress that looks great on one person looks terrible on another. The cut, the drape, the color against a particular skin tone — none of that reduces to a spec sheet. Every garment is a different product on every body. Anything where the value is entangled with the individual is beyond the reach of pure optimization.
Taste is yours. It evolves. It draws on context only your life can provide. No algorithm can decide what to buy on your behalf. Recommend — maybe. Decide — no.
There is also a raw infrastructure problem. No AI can query all the world’s information on the fly for every decision. The information keeps expanding, and costs grow exponentially as products have to be matched and judged against each other. Real markets — the kind that can actually be queried and compared — do not exist yet. They have to be constructed, opened and maintained.
Why Advertising Survives — But Changes
Even once those markets exist, advertising will not disappear. It will have to become honest.
An ad is a signal. It’s proof that a company is successful enough, or believes in its product enough, to fund the campaign. An AI shouldn’t ignore that. But the ad now has to be judged against the full market. It’s no longer about capturing attention or controlling visibility. The ad becomes data, and it has to earn its weight.
Here is the constraint that makes this stick: if an AI operates on any principle other than finding the genuinely best option for the person it serves — not best from a curated set of paid placements, but best from everything available — it loses trust. And trust is the only thing an AI agent has.
This is a Nash equilibrium. No player can improve their position by deviating from genuinely serving the user, because deviation means losing trust, and losing trust means losing everything. Advertising survives, but products now win on merit — not ad budgets.
The Resistance
Building real markets will not be easy. Companies whose monopolies run on psychology will never volunteer to be commoditized, and there are legitimate reasons to push back.
AI commerce today has real problems. Fraud is rampant — bad actors exploit automated purchasing. If an AI makes a buying mistake, the merchant eats the return costs. If an AI can scrape your catalog and feed it into a comparison engine, whatever product mythology you spent years constructing becomes transparent overnight. Merchants will restrict access, block bots, refuse to serve AI agents and lobby for legal protection.
This resistance is understandable. It will slow things down. But it will not stop them.
Every attempt to preserve information asymmetry against a technology that destroys information asymmetry has eventually failed. Companies that resist will lose incremental purchases and yield share to competitors who don’t. You cannot fight the incentives.
And technically, resistance will become impossible. When an AI agent can do exactly what a human does — run tools locally, use the same browsers, navigate the same interfaces — terms-of-service bans will be about as effective as drug prohibition has been at stopping people from taking drugs. People want to save money. People want easier lives. If they have the tools, no one telling them they can’t will succeed. The demand is too fundamental. The friction that sustained the old model was always just ignorance and inconvenience. Both are about to evaporate.
What Remains
Most of what we buy should not cost what we pay. When prices are set by real markets instead of by the apparatus of psychological manipulation that currently inflates them, GDP will shift. The fat gets cut. The real valuable work was never selling water bottles produced elsewhere for far less than we pay here. It’s the things we can do that nobody else can. The unique. The difficult. The genuinely creative. Work that can’t be commoditized because it isn’t interchangeable.
In his last shareholder letter as Amazon CEO, Jeff Bezos quoted Richard Dawkins: the universe is always pulling everything toward equilibrium — same temperature, same disorder, nothing distinguishable from anything else. The world wants you to be typical, Bezos wrote. It pulls at you in a thousand ways. You have to be different. In the age of AI, that is no longer advice. It is the only viable strategy.
What remains standing when the fog clears is what was always worth building. We are going to make money doing more interesting and more distinctive things. One struggles to see how that is anything other than progress.