Glossary
Judgment capital
There is an old test for whether something is capital, and it has nothing to do with how valuable the thing is. Judgment has always failed it.
Judgment capital (noun). Judgment that keeps producing a return when the person who formed it is not applying it. As distinct from judgment as labour, which returns only during the hours it is personally exercised.
The test
Labour earns while you work.
That line does the sorting everywhere else without any difficulty. A building earns overnight. A patent earns while its holder is on holiday. A brand earns in shops nobody from the company has ever visited. In each case the return has been separated from the presence of the person who created the thing.
Now run your own judgment through it.
Your judgment produced a return during the meeting where you spotted the flaw. During the ten minutes you spent rewriting the offer. In the moment you turned down the client everyone else wanted. Every one of those is real money and every one of them started when you started and stopped when you stopped.
Which makes it labour. Extraordinarily well compensated labour, frequently the best-paid labour in the building, and still labour by the only test that distinguishes the two.
Why this explains something that otherwise makes no sense
There is a pattern almost every founder recognises and few can account for. They own the company. The equity is theirs. By every conventional description they are an owner rather than an employee.
And they cannot stop.
Not because they lack discipline, and not because they have failed to delegate. The usual explanation is a personal failing of some kind, which is why the usual advice is about letting go, and why the advice reliably does not work.
The structural explanation is cleaner. The most valuable asset in the business was never an asset. It was a job, held by one person, that nobody had noticed was a job because the person doing it also owned the place. Everything else in the company had been converted into something that runs without its originator. Systems, brand, product, relationships. The judgment never was, because until recently there was no mechanism that could convert it.
So the founder who "owns" the business owns everything in it except the part that decides. That part is still a shift they turn up for.
The four properties judgment never had
| Property of capital | Judgment, historically | Once captured |
|---|---|---|
| Earns without the owner present | No. Stops when you stop | Yes. That is the conversion |
| Deployable in several places at once | No. One meeting at a time | Yes. Wherever the work runs |
| Compounds | Barely. Mostly re-earned | Yes. Each correction improves the next call |
| Survives the owner | No. Leaves with the person | Yes, if it was captured while they were there |
Look at the middle column. That is not a description of an underperforming asset. It is a description of something that was not an asset, and the fact that it was the most valuable thing in the business the whole time is the part worth sitting with.
The third row is where the distinction stops being academic. Labour does not compound: an hour of your judgment in 2019 makes an hour of your judgment in 2026 slightly better, through experience, and that is all it does. Captured judgment compounds in the ordinary sense, because each correction is kept and applies to work you have not done yet. Two identical operators, one capturing and one not, diverge on a curve rather than a line.
What this term is not, and the number I am not going to give you
Calling something capital invites a valuation, so here is the honest position before anyone offers you one.
There is no accepted method for valuing captured judgment. No comparable transactions to reference. No multiple, no standard, no line on any balance sheet that would survive an auditor. Anyone handing you a figure is producing it rather than measuring it, and the confident precision of the figure is the tell.
What is observable is directional and more useful than a number would be. Whether work you did not touch comes back the way you would have wanted. Whether that is happening more often this quarter than last. Whether the decisions reaching you are new ones rather than repeats of calls you have already made. Those are all things you can watch without inventing an asset class.
The term is also not a rename of anything. An Imprint is the store, the body of captured rulings. Judgment capital is what that store has become once it is returning something in work you are not doing. A store that exists and has never been put to work is an Imprint and is not yet capital. Same asset, two different conditions.
Where the conversion actually happens
Not at the moment of capture. That is a common misreading and it makes the whole thing sound easier than it is.
Capturing a ruling produces a stored ruling. Nothing has been converted yet, because nothing is earning. The conversion happens the first time that stored ruling governs a piece of work you were not part of, and produces the outcome you would have produced.
Which gives the term a fairly unromantic practical test. Point at a specific piece of work from last month. If you can say that the work was decided rather than merely produced, and that the deciding was yours, and that you were not there, something converted. If everything that carried your judgment last month also carried your attendance, nothing has converted yet, whatever is sitting in storage.
Frequently asked
What is judgment capital?
Judgment that goes on earning in work its owner is not doing. The distinction has nothing to do with how valuable the judgment is. It turns on whether the return depends on your presence, which is the line between capital and labour in every other context and has never been crossable here before.
Isn't a highly paid expert's judgment already capital?
No, and the size of the pay is what disguises it. Apply the ordinary test: does it earn when the expert stops? It does not. The return begins when they start working and ends when they stop, which is the definition of labour. A very high rate makes it excellent labour. It does not change the category.
How is this different from saying judgment is valuable?
Everybody agrees judgment is valuable and nothing follows from the agreement. Capital is a claim about structure rather than worth: it can be deployed in more than one place at once, it compounds, and it survives the owner's absence. Judgment historically had none of those properties, which is why the agreement never changed anyone's decisions.
Is judgment capital another name for an Imprint?
No, and keeping them apart matters. The Imprint is the thing: a body of rulings that has been captured. Judgment capital names a condition that body can be in, namely working and returning something without you. Rulings that have been captured and never put to use are an Imprint, and are not capital yet.
Can I put a number on it?
Not honestly, and the page does not try. There is no accepted method for valuing captured judgment, no comparable transactions to reference, and anyone offering you a figure is producing it rather than measuring it. What you can observe is directional and useful: whether work you did not touch comes back the way you would have wanted, and whether that is happening more often than it was.
Where this sits
The same asset argued on a different axis, why it was never priced, is at founder judgment. The store itself is at Imprint. The category is defined at What is Imprinted AI.
Related terms
- Founder judgment. The same asset argued on a different axis: why it never had a price.
- Imprint. The store. This term describes what the store becomes once it is working.
- Mirror Score. The nearest thing to a measurement, and its honest limits.
- The Assembler. The operator whose judgment stays labour no matter how good the tools get.
Where the term comes from
This is one entry in the vocabulary of a longer argument. The full glossary has fifteen terms. The case they belong to runs about 23,000 words, it is free, and there is no email gate on it.
Read The A.I. Business Manifesto
Nothing on this page is for sale. Quote it, argue with it, or pass it on.
Last updated: 28 July 2026