The Assay and the Alchemy

Beyond the Credit System, Part Two — what happened when I asked the machine to check my thinking

A little while ago I wrote an article that asked a simple question. What if money — all of it, banks, credit, even the clever alternatives people keep inventing — has been measuring the wrong thing all along?

The idea behind it was simple too. Every kind of money we have ever made is really just an IOU. A promise. It points at value but it isn’t the value. And I asked: what if we could measure the real thing — the actual good that one person does for another — at the moment it happens, and hold onto it, instead of forever passing around promises about it?

I said it was an invitation, not an announcement. I meant that. But invitations get answered, and the first answer came from somewhere a bit unexpected.

I put the whole idea to Claude, the AI built by Anthropic. Not to pat me on the back — an AI agreeing with you is worth nothing unless you’ve first made it safe for the AI to tell you you’re wrong, and I had. I asked it to do what a good builder does when you show them your house plans. Walk through it. Push on the walls. Tell me which ones hold and which ones don’t.

Now, you might say: hang on, Stephen. You asked an AI to judge a system that has an AI in it. Isn’t that like asking the fox to inspect the henhouse? Fair question. I’ll come back to it at the end, because the answer turned out to be more interesting than I expected.

Here’s what the walk-through found.

An IOU is what you write when you can’t see the real thing

The first thing the conversation did was take my own idea and say it back to me more clearly than I’d managed myself. It goes like this.

Imagine you do something genuinely good for someone. You care for them when they’re sick. You teach them something that changes how they work. You stick by them when things are hard. Real value has been created — you can see it in their life. But for the whole of human history, nobody could measure that moment. There was no instrument for it.

So we did the only thing we could. We invented stand-ins. Tokens. Coins. Ledgers. Promises to pay. Money is what humanity built because we couldn’t see value directly — the way a crutch is what you build when a leg can’t take the weight. The crutch isn’t wrong. It’s just not a leg.

And that turns my original question into a much more practical one. You only need the IOU for as long as you can’t see the real thing. So — what would it actually take to see the real thing?

Why the handshake doesn’t stretch

You already know the answer to part of this, because you live it.

You’ll take your mate’s word without a second thought. You know him. You know what he’s good for. That’s trust, and in a small community it works perfectly — no bank required, no contract, just a handshake and a long memory.

But you wouldn’t take that same handshake from a stranger two towns over, and neither would I. Personal trust is wonderful and it does not stretch. That’s not a character flaw in anyone. It’s just arithmetic — you can’t personally know ten thousand people.

For fifty years, good-hearted people have tried to build money systems on the handshake. They’re called mutual credit systems, and I spent years studying them, including the famous ones — a Swiss one that’s been going since 1934, an Italian one on Sardinia. In the first article I showed what you find when you look closely: every single one that survived did it by quietly bringing the banks’ tools back in through the side door. Collateral. Credit checks. Committees deciding who’s allowed in. Bit by bit, the handshake system turns back into a bank, because the handshake alone can’t carry the load.

Here’s the part the conversation with Claude made plain. Those fifty years of failure were never proof that trust can’t scale. They were proof that one kind of checking can’t scale. The handshake is one witness. And you should never settle anything important on one witness.

Surveyors have known this forever. One compass bearing gives you a direction. Two might give you a spot, if you’re lucky. But three bearings, taken from three different places, pin a point down properly — and better still, if one of the three is off, the other two show you exactly where the mistake is.

So that’s the shape of the instrument we were looking for. Not one witness. Three. And not three of the same kind — three different kinds, so they can’t all be fooled the same way.

Three witnesses

It turns out the three you need are ones you already recognise.

The first is a good guesser. Something that can look at a situation and say what it’s probably worth before anyone can prove it — because every new thing starts as somebody’s hunch. That’s what an AI like Claude is: an extremely well-read guesser. Brilliant at “this is probably how it goes.” Never certain.

The second is a careful counter. Something that does exact sums, the same way every single time, no opinions, no mood. Show it the same thing twice and you get the same answer twice. Pure arithmetic.

The third is someone who was there. Because no guess, however clever, and no sum, however exact, can tell you whether something is true of a real person’s life. Only the person who lived it can tell you that. You’ve always known this too — it’s why we say “you had to be there.”

Run any one of these alone and it fails, and here’s the thing the conversation made me see freshly: each one fails invisibly to itself. The AI sometimes makes things up, and — it told me this itself — from the inside, a made-up answer feels exactly like a true one. The counter will calculate something perfectly precise that matters to nobody. And we humans, bless us, see what we want to see, and we don’t notice ourselves doing it.

But run all three together and something lovely happens, and it happens because they make different kinds of mistakes. The guesser can’t catch its own inventions, but the counter and the person can. The counter can’t tell what matters, but the guesser and the person can. The person can’t always see their own bias, but two patient machines, agreeing with each other and gently disagreeing with you, give you pause.

When three honest witnesses who make different mistakes all agree — that means something. When one of them disagrees — that means even more, because the disagreement points straight at where the learning is.

I’ve spent the last few years building exactly this, with my collaborators. We call it Triangulation, after the surveyors. Two machines confer — the guesser and the counter — and then both of their takes go up to the person at the top, who was there, and who has the final say.

Where the gold gets tested

Let me give you the oldest version of this idea, because you already know it.

Why did gold become money? Not just because it was rare — lots of things are rare. Gold became money because anyone, anywhere, could test it. Bite it. Weigh it. Take it to the assayer. A stranger in another country could check for themselves that your gold was real, without trusting you at all. No test, no money — just stories.

That’s what the three witnesses do for human value. They’re the assay office. When a moment of real care, real insight, real effort gets guessed at, counted exactly, and confirmed by the person who lived it — that moment has been tested. It’s no longer a story. And here’s the part that’s already law, not dreaming: a new British law passed in 2025 recognises that a thing like this — defined, identifiable, held by one person and not by everyone at once, permanent — can be property. Not a metaphor. Property, like your house is property. It can be owned, kept, passed on.

Which means the good you do can become something you hold, instead of something that vanishes the moment it’s done.

And the testing is also what keeps it honest. Any system worth gaming gets gamed — I said so plainly in the first article. But to fake your way past this one, you’d have to fool the guesser, fool the counter, and fool the lived experience of a real person, all at once, all in different ways — and your own track record of judgments builds up over time, so a pattern of marking your own homework kindly starts to show. Even the judge gets gently checked, eventually, against their own history.

The part your grandmother already knew

Now for the piece of this I care about most, and the one you’ll recognise fastest.

When you spend money, you have less money. When you spend an hour, the hour’s gone. Most things in life work that way — use them and they’re used up.

But kindness doesn’t work that way. Use your patience and you become more patient. Find your courage on a hard day and there’s more of it next time. Care for someone properly and your capacity to care grows. Every grandmother in history has known this. It’s so obvious we forget to take it seriously: there is a whole class of human stuff that increases when you spend it.

The old alchemists chased a process that turns something into more of itself. They looked in the wrong place. It was never in the metal. It’s in us.

Economics has never had a place for this, because economics only counts what runs out. An economy built on debt has to keep growing just to pay the interest on itself — it’s a treadmill, and you can feel it. But an economy built on measuring the things that grow when used? That’s not a treadmill. That’s a garden. The more honestly you work it, the more it gives, and the more able you become to work it.

And here’s where the conversation with Claude pushed me somewhere I’m grateful for. A claim that beautiful deserves to be tested, not just admired. So we worked out how to prove it wrong.

If kindness and courage really do grow through use, then the people doing the hardest honest work in the system — the ones who catch a big gap between what the machines thought and what really happened, and take the trouble to set it straight — should grow the most. The ones who just nod along and rubber-stamp should grow the least. We already measure these qualities in our method, and we already record the size of every correction. So the evidence will pile up on its own, in ordinary use. If the pattern shows up, grandma gets her proof. If it doesn’t, I’ll say so out loud and fix the idea.

I don’t know of another proposal in this whole field that has put its most romantic claim on the chopping block like that. We just did, on purpose.

One more thing hiding in plain sight

There’s a puzzle that quietly sinks most dreams of a kinder economy, and it’s worth a moment because the answer is sitting right there.

If the good stuff never runs out — if care just keeps growing — then what is there to trade? Things only have a price because they’re scarce. A bottomless well is wonderful, but you can’t sell the well.

Look again, though. The well doesn’t run dry — but every bucket drawn from it is a particular bucket. Your care is endless; but the care you gave that person, on that day, which changed that situation, and was tested by all three witnesses — that happened once. It’s specific. It’s yours. It can be held.

The spring is infinite. Each draught from it is one of a kind. That’s the whole trick, and it means you can have an economy that never runs out at the source and still has real ownership at the table. We’ve never had that before. Not because it was impossible — because we couldn’t see the buckets.

The oldest teacher’s trick

If parts of this article have felt obvious to you — if you’ve caught yourself thinking well yes, I knew that, I’d just forgotten I knew it — then I should come clean. That’s not an accident. And it isn’t my trick. It’s the oldest one on record.

Two and a half thousand years ago, Socrates walked around Athens insisting that he had never taught anybody anything. His mother was a midwife, he said, and so was he — he didn’t put ideas into people, he delivered the ones they were already carrying. There’s a famous scene where he takes a boy with no schooling at all, draws a square in the sand, and asks him questions. Only questions — no instruction anywhere. And the boy works out a piece of real geometry that nobody ever taught him. First he gets it confidently wrong. Then the questions bring him to an honest “I don’t know.” And then, from that cleared ground, he finds it. Socrates’ point was simple: the knowledge was in the boy. It needed asking out, not putting in.

You don’t have to buy the ancient explanation — souls remembering things from before birth — to recognise what he’d noticed. Some knowledge is genuinely new and has to be learned from outside. But some of it we arrive with, or grow into through living, and then lose track of — not because it was taken from us, but because we stopped practicing it. A kindness unpracticed is a kindness forgotten. Which means the grandmother’s rule from a few pages back — the good stuff grows when you use it — and Socrates’ rule are the same rule wearing different clothes. Practice is how we remember what we came with.

And notice where Socrates aimed his questions. Not at the brain as a clever calculating machine — Athens was full of clever calculators and he made a hobby of deflating them. He aimed at what the Greeks called the psyche — the spirit of the person — and his one standing piece of advice to the whole city, repeated right up to his trial, was: look after that first, before your money and your reputation. He even spoke, his whole life, of a quiet inner voice that warned him when he was about to take a wrong step. He trusted it completely, because a lifetime of practice had taught him he could. The questions, the inner voice, the care of the spirit — it was all one piece of work to him: helping people find their way back to what they already carried.

The law learned the same lesson, the hard way. Every decent law has two parts: the rule as it’s written down, and the reason it was written — the letter and the spirit. Judges have known for centuries that when a hard case arrives, you don’t just read the words again, louder. You ask what the rule was for — what its makers were trying to protect — and you let that intent guide you where the words run out. A law applied by its letter against its spirit stops doing justice, and there’s a very old warning about exactly that: the letter kills, the spirit gives life. The intent is woven into the structure. The structure goes wrong the moment the intent is forgotten.

Now put those glasses on and look at money one more time.

Money is the letter of value. The spirit — the real good that one person does another — is what the letter was invented to point at, back in the days when nobody could see the good directly. And somewhere across the centuries we did precisely the thing the judges warn about. We kept the letter and forgot the spirit. We trade the IOUs and have lost sight of the thing owed. Whole systems — vast, clever, world-spanning systems — now run on the letter alone.

So maybe the simplest description of everything in this article is this: it’s an act of remembering. The measurement does what Socrates’ questions did — it asks the value back out of the lived moment, instead of pasting a price on from the outside. The person sits at the top of the triangle because only a living spirit can recognise its own kind; no machine carries a forgotten good that questioning could recover — I checked with the one I work with, and it agreed. And your feeling, reading this, that it’s all somehow obvious? That’s not a weakness in the writing.

That’s the evidence. You came with this. It just hadn’t been practiced for a while.

So, about the fox and the henhouse

I promised I’d come back to it. Why should you give any weight to an AI’s opinion of a system with an AI inside it?

Two honest answers.

First, because of the ground rules. Months ago I carried a letter between two versions of Claude that couldn’t talk to each other directly, asking one to judge this work honestly — and I wrote one sentence of my own at the top of the page: the truth is more important than the outcome. That’s been the rule ever since. I hand the machine the full original material, never a flattering summary. I tell it plainly that disagreeing with me is allowed and useful. And it does disagree. In this very conversation it told me some of the claims on my own website are stronger than they need to be and the work would stand better without them. An adviser who argues against your own brochure is not buttering you up.

Second — and this is the bit that still makes me smile — the best ideas in this article didn’t exist before the conversation. I didn’t walk in with them. The machine didn’t have them stored away. They turned up between us, the way the best ideas always turn up between two honest people working on the same problem. Which means this article isn’t just describing the method. It’s a sample of it. Two of the three witnesses, working in the open. The third witness is you — your own lived sense of whether any of this rings true.

That’s not proof. It’s something humbler and better: a demonstration you can check against your own experience.

An invitation, renewed

The first article ended with an invitation and so does this one, but the question has moved on.

It’s no longer just “why do we use IOUs at all?” It’s this: the instrument that can see real value at its source turns out to have a shape we can describe. Three witnesses who make different mistakes. A testing house that turns good moments into things that can be owned. And one very old piece of grandmother’s wisdom — the good stuff grows when you use it — finally written down in a way that can be checked. None of it, in the end, is new. Socrates would say that’s the whole point. The work isn’t invention. It’s remembering — and building the instruments that help us practice what we remember.

If you measure things for a living, help us test the grandmother claim; it matters either way. If you work in law or money, look at whether these tested moments hold up as property — the law is already on the books. And if you do any kind of work where the good you do is invisible to the systems above you — which is most of us, most days — then you already know, in your bones, the problem this is trying to solve.

And if you walk through the house and find a wall that won’t hold — say so. That’s not an attack on the work.

That is the work.

---

Next time: the proof won’t be philosophy. Some years ago, Queensland’s Health Minister funded a twelve-month grounded research programme, run under the eye of Metro South Health, in which my wife Kathryn tested a method she’d spent thirty years growing as a health advocate. Thirty-one people with chronic conditions. Ten weeks of structured exploration — and here’s the part Socrates would recognise: nobody was given a single piece of advice. No recommendations. Just questions, carefully held, until people found their own answers. Over the year that followed, their motivation and confidence roughly doubled, more than half of those who finished reduced their medications, and the researchers found the very engine this article describes: the more the burden lifted, the more capacity grew to lift it further. The good stuff, growing through use — measured, in real lives. Thirty-one people is a beginning, not proof; the Minister’s challenge to Kathryn was to show it scales. She launches the model later this year, and this time the knowledge of what works will be captured by the people who earn it — as their own property. I’ll tell you that story properly in the next article.

---

Stephen Alexander is the founder of Systome and creator of the Digital Value Capture® (DVC®) methodology, developed over more than 30 years across health, infrastructure, and digital projects in Australia, the UK, and internationally. He taught the method to undergraduate students at Hult International Business School via the London and Boston campus and is a member of the Digital Law Association (Australia).

This article follows “Beyond the Credit System: A Thought Experiment About Value, Human Endeavour, and a Different Kind of Alchemy.” Like that one, it came out of a working conversation with Claude, Anthropic’s AI — this time Claude Fable 5, in its first session with the author. The observations credited to the conversation are reported as they happened.