Ø Frozen Predictions · the document of record

Pre-registration: α(0) does not vary across the record history

Freeze date: 2026-09-11 Author: G · Studio G · the420code.org Scope: AP51 (The Glass), FINAL v1.0, §4 and §10; carried by KS-GLASS.3. Status: frozen in the wave of the four, at the author’s word. The commit hash is the timestamp: c9c02cd, tagged wave-2026-09-11.


1. The claim

The fine-structure constant at zero momentum transfer, α(0), compared under equivalent conditions, has no intrinsic variation across the record history — not with place, not with epoch, not with the density of the landscape. Every measurement of Δα/α returns a value consistent with zero within its uncertainty; individual estimates scatter, and the prediction is that no variation is confirmed at any precision.

2. Scope

Spatial and temporal variation of α(0). Not the running of the effective electromagnetic coupling with momentum transfer, which is AP50 §10’s question and D-EM.1’s.

3. Predicted

Δα/α = 0, structurally. No number of the corpus’s own. There is nothing here to fit and nothing to tune: the prediction is the absence of an effect.

4. Measured

Oklo; quasar absorption lines; atomic-clock comparisons. Contested detections exist — the spatial dipole claimed from quasar absorption in 2011 — attributed to systematics; the newest measurements return zero.

5. The kill condition

A variation of α(0) reported at five standard deviations or better, by two independent methods, standing after their systematics are reviewed. Then P-GLASS.1 is dead and shown; KS-GLASS.3 fires; AP51 §4’s universality is withdrawn; AP50 §3’s sorting stands or falls on its own.

6. What a null does not do

It does not single the hourglass out: every constant-α picture shares the prediction. What the hourglass adds is the reason — the grain is the same at every site and every tick, and the constraints are its readings. A null confirms nothing on its own. It is registered because a prediction that cannot distinguish must still be written down before the measurement, or the record is incomplete.

7. Standing, and what is not yet true of it

Frozen11 September 2026, in the wave of the four, at the author’s word
Commit that timestamps itc9c02cd, tagged wave-2026-09-11
Document of recordAP51 The Glass, FINAL v1.0, §4 and §10
Kill switchKS-GLASS.3 — LIVE — STRUCTURAL; EMPIRICAL through P-GLASS.1
Independent anchoringOutstanding. See below.

On the anchoring, stated plainly. The freeze bundle for this wave was prepared for the independent verifier on 11 September 2026. As of 12 September 2026 the verifier’s own anchored proof repository carries nothing later than the AP49 bundle of 6 September, and his published pre-registration pages do not name P-GLASS.1. This registration therefore stands on the author’s word and on the commit that timestamps it, and not yet on an independent countersignature. When the anchoring lands it is recorded here, dated, and this paragraph is superseded rather than removed.

Frozen files are never edited; corrections supersede, dated, in the open.