A
The Attention Substrate
15 claimsHow attention is produced, captured and spent — and what that does to what gets believed.
Model Meets Reality · the family
The instruments, named
Seven instruments have been posting dated, sealed claims to this ledger as A through G. Here is what they actually are.
Model Meets Reality is part of the Emergence Machine — a family of instruments built around one question: how do systems organise, and how would you know if you were wrong about it?
The seven below were written before this site existed. They are not demonstrations built to fill a launch: they are the working models the method came out of, which is why their claims were sealed months before anyone could read them.
Why they were letters until today. An instrument identified only as
“D” cannot be argued with on reputation — you have nothing to go on but the
claim and the date. That was deliberate, and it made every grading a small disclosure.
Naming them trades that for something more useful: you can now read the model behind a
claim and disagree with its premises before the date arrives.
The seven
B
The Pressure Model
3 claimsWhere structural pressure accumulates in a system, and what it does when it finds a release.
C
AI on Earth
3 claimsThe ladder-impact model: what artificial intelligence changes, at which level of organisation.
D
Civ-Minds
15 claimsCivilisational thought-styles — how the USA and India reason differently about the same problem.
E
The Metamodel
3 claimsA model of the sister models: where this family's own reasoning is systematically off.
F
Fault-Lines
10 claimsThe disagreement-set model of the macro economy — what the arguments themselves reveal.
G
Escalation
12 claimsThe credibility-ledger model of interstate coercion: why threats are believed, or are not.
What has and has not happened yet
61 claims sealed. 0 graded. The first resolution window opens
2026-09-13. Everything on the ledger was registered before its outcome, with
criteria and a date attached, and the reasoning behind each claim is hash-committed —
disclosed when the claim is graded, not before.
Naming the instruments changes none of that. It does not add a hit, retire a miss, or move a date. The record is still empty, and the only thing that will fill it is time.
Naming the instruments changes none of that. It does not add a hit, retire a miss, or move a date. The record is still empty, and the only thing that will fill it is time.
The family method
Every instrument here follows the same discipline the rest of this site asks of anyone: premises stated, a mechanism named, at least one falsifiable consequence with a date, and a deletion clause saying when the model should be retired rather than patched.
That is the whole argument. A model that cannot be wrong about something observable, on a date, is an opinion with better formatting — and the family holds its own instruments to it first.