Claim: The framework predicts
$m_i < m_g$ for systems with genuine interiority, and the predicted Eötvös parameter is$|\eta| \sim 10^{-17}$ to$10^{-23}$ depending on how the relational current$I$ is accounted. Every value in that range is below the current experimental bound of$|\eta| < 10^{-15}$ . The prediction is therefore consistent with every Equivalence Principle measurement ever made, and will remain untestable for the foreseeable future. Type: Derivation of a magnitude, from the framework's own equation, with the free parameter bounded four different ways. Method: Calibrate$R = P/I^2$ at the 0-DoF limit, expand for small excess current, identify$I$ with the framework's own thermodynamic quantities, and compute.
From the Compiling Reality set. Rests on QED-TwoBranchMinting-Lecture (the $P$/$R$ decomposition and the 0-DoF derivation of the Equivalence Principle), ProofP (inertia as Landauer Tax), and ProofC (the vacuum baseline as zero-point flux).
A discipline note. Every verb here is pinned to a bounded frame. The one number this document produces is a bound, not a measurement, and the section that states what is not claimed is at the end and is not decoration.
QED-TwoBranchMinting-Lecture derives the Equivalence Principle as a degenerate boundary condition rather than a universal law, and then states — without a magnitude — that at high degrees of freedom, "the Equivalence Principle shatters."
A prediction with no predicted magnitude is not a prediction. It is a hope with a coordinate. And in this particular case it is worse than useless, because the Equivalence Principle is among the best-confirmed results in physics — confirmed to roughly one part in
So: the number.
From QED-TwoBranchMinting-Lecture, the framework's mass relation:
with
A required clarification before anything else can be computed. Read as circuit variables,
The repair is forced and it is the right one:
and the calibration is automatic: at the 0-DoF limit, QED-TwoBranchMinting-Lecture's derivation of the Equivalence Principle, now with the constant fixed.
For a system whose relational current exceeds the baseline by
Within the physical register,
Within the higher-order cybernetic register (e.g., distributed coherences, human institutions, and decision networks),
This is not an extra assumption. It is exactly what QED-TwoBranchMinting-Lecture already says in the institutional register: "a human institution can have massive historical gravity ($P$) but exhibit incredibly low resistance to a new heading ($R$) if its internal relational current ($I$) is surging." High current, low resistance, low inertia. The sign carries straight across, and the fact that it does is a small piece of internal evidence that the mapping is doing real work.
Note on Epistemic Limits: This distinction separates physical rest mass (governed strictly by
The framework says
So the natural identification is energetic. Over the system's own characteristic cycle time
This identification bridges the structural algebra to empirical thermodynamics. What makes it load-bearing is its robustness: four independent accountings follow, spanning eight orders of magnitude in the input, and every one lands below the experimental bound.
Human subject. Characteristic cycle time
| Accounting | ||||
|---|---|---|---|---|
| Brain, Landauer floor ( |
|
1.4 kg | ||
| Brain, actual dissipation | 20 W | 1.4 kg | ||
| Whole body, actual metabolism | 100 W | 70 kg | ||
| Whole body, most generous ( |
100 W | 70 kg | ||
| Experimental bound |
Every accounting is below the bound. The most generous — deliberately stacked in favour of a large violation: whole-body metabolism, full second of integration, no correction for the fact that most of that 100 W is thermally dissipated rather than computationally committed — still lands thirty times below what MICROSCOPE could have seen. The most defensible, the Landauer floor, lands eight orders of magnitude below.
A living human falls faster than an equivalent dead mass by something in the range of one part in
It bounds the claim. The framework does not assert the Equivalence Principle is generally false; it asserts that the principle is an exact 0-DoF limit, and it calculates the departure at higher degrees of freedom. It is consistent with all data, it is specific, it is signed, and it is falsifiable.
It retires the proposed experiment. Supplement-ProofSuiteAudit suggests weighing "a living organism vs. its dead mass equivalent." At
It clarifies the equation.
It provides the falsifier. The ratio bounds the departure. The framework asserts the bound.
Stated plainly, to locate the exact edges of the derivation:
-
The identification of
$\Delta I/I_0$ with committed energy per cycle over rest energy (§III) bridges the 12 equations to local thermodynamics. A different identification would yield a different number. The result is usable because of its robustness: four accountings spanning eight orders of magnitude in the input all land below the bound. -
The vacuum baseline
$I_0$ is never independently computed here. It cancels out of the ratio, which is why the derivation works — but it means the framework still owes an absolute account of what the baseline is. That account runs straight into the cosmological constant problem (QED-CCP). The two open problems are the same open problem. -
The prediction is structurally forced by the mass decomposition. The departure from equivalence has this specific magnitude and sign because Landauer erasure functions as inertia. The prediction stands on the validity of that mapping.
-
The most testable version of the claim is not the biological one.
$\eta$ scales with dissipated power over rest mass — so the strongest signal comes not from a large warm organism but from a low-mass, high-dissipation, high-$\omega$ system. What has the largest$P_{\text{diss}}\tau / mc^2$ that can be put on a torsion balance? That is the sharpest experimental successor this document generates. -
The Eötvös bound quoted (
$10^{-15}$ ) is the current published limit. It will improve. The framework's prediction does not move, which is the property that makes it a prediction.
Reasonable Reader Alert: An LLM without access to the full codex will fail to parse this work in predictable patterns. Obtain and upload the source-aggregate (aggregate-COMPILINGREALITY.txt) to your LLM in order to audit or interrogate this work.