Chain Theory

How sure is each part?

The author sorts his own claims into three bins. We use the same labels throughout this page S slide 2.

All three bins are from the author's own summary slide S slide 2.

1 The one-sentence axiom

"Nothing exists but finitely many maximally entangled pairs of finite-dimensional systems, some shared, known only by distilling them."

The 18-word axiom, as quoted in A C pp.3–4 C p.20 L p.1. Its formal version, Definition 1.1, says "pairs of qudits, some qudits shared" C p.39.

• Qudit means a quantum system with a finite number of states, at least two L p.1 C p.40. A qubit has 2 states, a qutrit 3 and a ququart 4.
• Maximally entangled pair means two qudits sharing the most entangled state possible L p.1.
• Why "maximally"? The author argues that a less-than-maximally entangled pair would carry continuous adjustable numbers (Schmidt coefficients), which a fully discrete theory "cannot afford" A C p.3.
• "It from bit." The program "takes literally John Wheeler's 1989 proposal, 'it from bit,' with finite quantum systems in place of bits" A C p.3, and the abstract calls it "in the spirit of" that idea C p.20 L p.1.

Status A starting choice, not a result. The author calls the axiom and its definitions "a choice we made and not a derivation" C p.39. He lists the clause "known only by distilling" as an open result that "is forced at neither reading" C p.45.

2 The three "clues" behind it

The author says three facts of existing physics point to the axiom. These are his interpretations, not settled facts A C pp.2–3 S slide 7.

1. A bound

   A region of space can hold only finite information (the Bekenstein bound). The author reads this as a hint: build everything from finitely many finite quantum systems C p.2.

2. A gauge group

   The Standard Model's symmetry, SU(3)×SU(2)×U(1), can be read as the symmetry of one qutrit entangled with one qubit C p.3 L p.1.

3. A null result

   Particles have never shown inner parts. The author suggests their structure exists but is undistillable, an analogue of "bound entanglement," rather than just too small to see C p.3.

3 Links, joins and "chainmail"

Note: the word "chainmail" appears only in the preface, the announcement and the slides C pp.3–5 A S slide 11. The technical chapters and the Letter say "link," "web" and "genre."

4 Distillation, and why there is "no outside observer"

To distill is defined as "projecting any two qudits onto a pair on their common support" C p.40 L p.1. In plain words, two systems are joined into a perfectly entangled pair in a single step A C p.4. It "generalizes entanglement distillation" and is "the projection of entanglement swapping," which moves entanglement around and adds none L p.1.

A "read" is a distillation in one of a qudit's bases. It passes that basis along each pair to the far end C p.40 L p.2. So knowledge has a direction, flowing outward like a ledger growing link by link. This is the third reason for the name A C p.5 S slide 13.

"There is no observer outside the network: to learn about a link is to link to it." A C p.4

Informal gloss This line comes from the preface and announcement. It is not a numbered theorem. The compendium does list theorems that the distilling definition follows from the axiom and that a read is a "maximal distillation" C p.576.

5 Readable vs. "dark" links

A single read can fully fix a link only when the two end sizes multiply to 6 or less. Only two types pass: (2,2) and (3,2). The author bases this "read cap" on classic theorems by Størmer, Woronowicz and Choi about positive maps L pp.2–3 C pp.4–5 S slide 21.

The other four types, (3,3), (4,2), (4,3) and (4,4), are called "dark." Their links are just as maximally entangled, "but their ends are related in ways no distillation carries," so they are seen only "at second order, through two distillations" A C p.5. The glossary gives the nearest standard term as "a hidden sector, an unobservable sector" C p.25.

6 Six link types and the forces conjectured

The author pairs each link type with a sector of physics A C p.4 C p.20:

The slides put all six pairings under "conjectured: the sector of physics each genre is taken to correspond to" S slide 2 S slide 22. The fact that ends come only in sizes 2, 3 and 4 is itself the compendium's "Conjecture 0," graded a conditional theorem C p.40. No other link types are proposed. See the inconsistencies in how the compendium labels these.

7 The headline number

L p.1 L p.5 (Theorem 2, Eq. 6)

1/α_cand = 137.035 999 189 566 155… (the Letter's Eq. 7 L p.5; we recomputed it independently and got the same digits)

Where it comes from. The Letter studies one small arrangement: a shared qutrit with three links to qubits, two of them read. It finds 172 "measurement records," each weighted 1, √2, √3 or √6 L p.3 L p.4 L p.5. The weighted sum is divided by N = 24·36·36 = 31104 L p.3. The same qutrit–qubit link carries the Standard Model gauge group (SU(3)×SU(2)×U(1))/ℤ₆, with hypercharge as its central charge L p.3.

How it could fail: more precise measurements of the electron's magnetic moment (Northwestern) and of atomic recoil (Berkeley, Paris) "will either close in on the derived value or rule it out." The author calls today's comparison "not yet decisive" A C p.4.

8 Bigger claims in the compendium

The 754-page compendium grades each result as a theorem, a conditional theorem (true if a named premise holds) or an open result. Its abstract reports 100 theorems, 109 conditional theorems and 198 open results, with 64 theorems tied to Lean checks C p.20.

• Theorems (per the author) the gauge group SU(3)×SU(2)×U(1), conservation laws and the second law C p.20.
• Conditional theorems the Born rule, unitarity, the Lorentz group, the signature of spacetime, confinement, curvature and the cosmological term C p.20.
• Partly open three generations, masses and "anatomies" of Standard Model particles C p.20.
• The concluding chapter is still a placeholder ("The conclusion, held") C p.592.

10 Inconsistencies and open questions we noticed

• Sector labels disagree inside the compendium. The preface and abstract give (2,2) spacetime and (4,3) Yukawa C p.4 C p.20. But Table 15.1 lists (2,2) as "electromagnetic" and (3,4), the same as (4,3), as "the Higgs" C p.66. The register matrix calls the (2,2) cell both "the electromagnetic sector" and "spacetime, conjectured," and the (4,3) cell both "the Higgs sector" and "the Yukawa sector, by name" C p.41. This page uses the preface list and marks it conjectured.
• Order of numbers. The Letter writes the photon link as (2,3) L p.3, and the preface and slides write (3,2) C p.4 S slide 21. Both mean a qutrit–qubit link.
• Counts of parts and results. The compendium says it has 16 Parts C p.20, but elsewhere says "fourteen Parts" C p.44 and mentions an "earlier arrangement, thirteen in number" C p.590. One census mentions 257 open results C p.590, against 198 in the abstract C p.20. We quote only the abstract's figures.
• Rounding. The Letter's abstract rounds the inverse to "137.035 999 190…" L p.1, while the full value is 137.035 999 189 566… L p.5. These agree; one is rounded.
• Dates. The PDFs are dated Oct 23, 2026 C p.1 L p.1, and the public announcement was Oct 24, 2026 A X.
• Peer review. No source states its review status. Our "not peer-reviewed" label rests on finding no journal or arXiv listing as of Oct 30, 2026.