The Coin in the Drawer

Nothing malfunctioned that night. The feed is a price system, and it did what price systems with exactly one price have always done. There is a name for the pattern, the name is four and a half centuries old, and it comes with a prediction we are currently living inside.

Half a law

In 405 BC, Aristophanes put it in a chorus. In The Frogs, he notes that Athens’ old silver coins — full-weight, trusted around the Aegean — had disappeared from the market stalls, while the shabby emergency bronzes struck the year before passed busily from hand to hand. The city, he said, was handling its best citizens the same way. Nicole Oresme worked the pattern out properly in the fourteenth century. Copernicus wrote it down in 1526, in a treatise on coinage composed for a king. And in 1858 the economist Henry Dunning Macleod pinned it on Sir Thomas Gresham, the English Crown’s money agent, whose actual contribution had been to warn Elizabeth I about debased coin. The law about counterfeit things circulating under an honest name is itself named for the wrong man. That feels about right.

Bad money drives out good. Anyone can recite it. Almost nobody keeps the clause that does all the work: only where both coins are forced to trade at the same price. The engine of the law is the par — a legal-tender rule that says a clipped coin settles a debt exactly as well as a full one. Hand people that rule and they need no instruction in what to do next. They spend the bad coin and keep the good one. The good coins were not destroyed; they were hoarded, melted, carried abroad — into drawers, under floorboards, across borders to wherever their weight would be priced again. And where the rate floats, where a full coin openly buys more than a clipped one, the law runs backwards and good money drives out bad.

The famous sentence is not a law about money. It is a law about what a forced equality does to unequal things.

Figure 01

The par is the engine

Fixed par — one price for every coin

• Clipped coin → spent → circulates everywhere
• Full-weight coin → held back → drawer · melt · export

Floating rate — each coin at its price

• Clipped coin → discounted → declines or dies
• Full-weight coin → premium → circulates openly

The feed is the par

AI did not debase writing. It did something more interesting: it made minting free. Every model is a mint with no metal cost, striking as many coins an hour as anyone cares to ask for. But free minting alone does not trigger Gresham’s law — cheap mints and forgers are ancient. The law needs the par, and the par is the feed. One slot, one card, one autoplay window, one swipe. The feed’s founding promise — anyone’s post, same surface — is a legal-tender law for attention. It prices placement, so it cannot price effort. A paragraph that cost an evening and a paragraph that cost a twentieth of a cent settle the same debt there: one glance, at par.

So the law fires exactly as written, and the symptom is the one the drawer always shows. The open web is not so much filling with synthetic text as it is emptying of the other kind. The writers are not quitting. They are hoarding. The serious essay goes to the small list. The real argument happens in the group chat. The piece someone would stake a name on circulates in rooms with doors, person to person, where the writer’s weight gets priced again. We keep reading this as decline — nobody writes anymore, the internet is dead — when it is the oldest monetary behavior on record: the good coin going home.

And the instinct everyone reaches for — detection — is the fix that has never once worked. Assaying every coin at the till costs more than the trade is worth; that is why legal tender exists in the first place. No currency was ever saved by better bite-tests. They were saved by letting prices differ.

Five predictions

1. The open feed stops pretending.

   The general feed becomes the bronze tier and settles into the role. Most of what circulates at par will be machine-struck, and most people will not mind, because nobody who holds a good coin still spends it there. The scandal fades into furniture, the way nobody resents small change for being zinc.

2. The drawer becomes the first-run market.

   Group chats, small lists, and invitation rooms carry a growing share of the writing people actually act on. “Published” splits into two words: issued, which means struck and released at par, and circulating, which means passed along by named people who priced it first. Prestige follows the second.

3. Some platform floats the rate — out of inventory, not ethics.

   An all-par feed eventually holds no coin worth collecting, and a feed nobody saves is a business problem. The first major platform to price provenance and effort differently — a lane, a mark, a separate surface that a machine cannot buy its way onto — pulls hoarded writing back into the open, and gets copied within a year of visibly working.

4. The assay moves from the coin to the mint.

   Machine-text detection keeps losing, because the metal is now identical. So the stamp migrates from the artifact to the author: persistent identity, staked reputation, a person who answers for the words. The question stops being “was this generated?” — unanswerable — and becomes “who answers for it?”, which always has an answer.

5. The measurements go wrong in one direction.

   Everything that reads culture off the public web — trend reports, research corpora, the general sense of what people think — starts reading the hoard as absence. An empty till was never a silver shortage. Institutions that treat the feed as the whole economy will be late, together, in the same direction.

At par, daily

I publish here every day, into the open, at the feed’s price. I have already argued that most of what reads the open web is no longer a person, and that attention is leaving the economy that was built to harvest it. This piece is the third leg of that stool: where the valuable writing goes while all that happens. I keep striking coins at par on purpose — a blog is a mint, and a mint that stops issuing tells you nothing about the metal. But I notice where the writing that changes my mind arrives: in small rooms, from named people, at a rate we set ourselves. The stamp I trust is the person, not the coin — and the counterfeit I watch for rings the bell without moving the weight.

The claim ledger

Already true

Minting costs approximately nothing. Detection keeps failing in public. Serious writing has been sliding toward newsletters, chats, and closed rooms for years. Feeds price placement, not provenance — the par is already law.

What has to happen

A major platform visibly prices provenance or effort differently — a lane, a mark, a separate unit — and hoarded writing measurably returns to it. Watch for the first one that works, and for the imitators inside a year.

Where I am probably wrong

Ranking may already be the floating rate. If recommender systems quietly price the dear coin higher, reader by reader, then the par is a surface illusion, the drawer stops deepening, and this essay describes the recent past rather than the future. The law needs a true par — and the algorithms may have already repealed it without telling anyone.

It is late on a Tuesday, and nine people are reading what you wrote. You did not stop writing this year. You stopped spending it at par — which is not retreat, but the oldest judgment a holder of value knows how to make. The drawer was never where the coin went to die. It is where the coin waited for a market that could tell the difference.

Yours still weighs something. Mind where you spend it.