# Dated note beside AP12 The Limit — issued 12 September 2026

**The 420 Code · Studio G · a supersession note; the frozen file is never edited**

AP12 (*The Limit*, Uncertainty, Notebook III) is published at
[the420code.org/AP12_The_Limit.pdf](/AP12_The_Limit.pdf), 22 pages,
sha256 `396c2583061320b3f5adadd49ee4bb17c761ec165eaf7856f676b58b9670c5a6`.
One matter of record has arisen on §1.3, the five-step forcing argument for ℏ.
**It does not alter the body of the paper, and the paper is not edited.**

> 12 September 2026, on §1.3. Step 1 misstates Stone’s theorem: the theorem fixes the generator K
> of a strongly continuous one-parameter unitary group with a given time parameter, U(t) = e^(−itK);
> it does not fix a separate action constant, since H = aK gives the same evolution for every
> positive a. Step 4 relies on a minimum non-trivial unitary evolution; a strongly continuous group
> has none — U(t) exists for every real t — and discrete outcomes do not discretise the evolution
> between them. **Both steps are withdrawn.**
>
> What stands is Step 2 — the axioms carry exactly one action scale, B × R — together with the
> identification that one record carries it, which is AP46 §5’s tick by construction: one grain,
> one tick, one ℏ. **§1.3 is therefore regraded from DERIVATION to IDENTIFICATION on the author’s
> rule, by construction.** KS-Q.8 is closed on that rule and not on the five steps; KS-L.1’s status
> line reads so. KS-L.4 stays LIVE; Step 3’s inventory of the axioms’ dimensionless numbers (2 and
> 1) is superseded by AP28’s and AP50’s counts (3, 4, 6, 18, 21, and 1/π), none of which enters the
> action scale.
>
> §2 and §3 — the commutation relation and the Mandelstam–Tamm bound — are untouched: they use ℏ as
> the scale and never lean on the forcing. Claim Summary §1 reads IDENTIFICATION on the author’s
> rule.
>
> Found in review by a panel of AI models, 12 September 2026; verified against the PDF of record,
> locus by locus.

**Grade.** BLOCKING for KS-L.1’s status line. Nothing else in the paper moves. The uncertainty
principle, the commutation relation and the energy–time bound are unaffected; ℏ as the action of one
record is unaffected. What is withdrawn is the claim that the axioms *force* it rather than
*identify* it.

## What the registry still says, and what is owed

The Master Kill Switch Registry v5.31 carries this switch as `KS-Q.8 / KS-L.1`, CLOSED, with its
ground given as “Stone’s theorem + Axiom B forcing argument”. That ground is the part withdrawn
above. The switch stays CLOSED — it closes on the author’s rule, by construction — but the
registry’s stated reason is superseded by this note. **The registry line itself is owed at v5.32**;
it is not rewritten here, because the room that renders the registry renders it verbatim, and a
laundered row would disagree with the PDF served beside it. The room carries this note against the
row instead.

## Why this note is beside the paper and not inside it

No source of record exists for AP12: the closest candidate file differs from the published text, as
recorded in `SOURCE_OF_RECORD.md` in this folder. Until a matching source is produced the published
PDF is the source, and the republish procedure that places a note *inside* a paper is not available
to it. The note therefore stands beside the paper, linked from its row on Ø Proofs.

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