Proposal note · shared by link only
One research programme, two doctorates — a proposal
1One programme, not two
G's lane — executable falsifiability
G published an axiomatic corpus (the 420 Code — Artist's Proofs and Ø Predictions, DOI 10.5281/zenodo.19208226, CC BY 4.0) that does something most theories never do: it indexes its own refutation. The corpus publishes 549 falsification conditions ("kill switches"), 522 of them individually named, and makes five headline predictions from a single measured input (α, the fine-structure constant) with zero fitted parameters — each prediction carrying an explicit, published condition under which it is dead.
The doctoral claim this proposal carries forward from this lane is the method, not the physics: Popperian falsification made operational — refutation conditions that are named, typed, and executable, so that "falsifiable" stops being a philosophical adjective and becomes a property a machine can check. The natural home for that claim is the philosophy of science.
AJ's lane — self-verifying systems
The CongoSky estate is built on one house convention, stated in the proofs directory itself: each proof is "runnable, dependency-light, and self-verifying — it exits non-zero if any claim fails." Dozens of proof directories back the estate's claims with working code instead of adjectives — among them the two physics receipts linked in §4, the One Record store, and a cite-or-refuse answer engine that quotes a source or refuses, live at klaus.congosky.cloud.
Two lanes, one discipline: a claim you cannot execute is a claim you have not yet made. That shared discipline is why this is one programme and not two repositories stapled together — the spine below is the load-bearing argument.
2The coherence spine
The line that braids the two bodies of work runs unbroken from G's axiom to the estate's safety machinery:
Axiom R — reality as irreversible, append-only records
↓ instantiated as a storage engine
One Record — append-only, delta-only, content-addressed, Merkle-chained
↓ answered over, honestly
cite-or-refuse — the AI that would rather say nothing than lie
↓ closed by the same discipline
kill switches — falsification conditions as executable Popper
G contributes the axiom and the falsification discipline; AJ contributes the executable instantiation and the systems evidence. One spine, two proposed theses.
Salt on the spine
The axiom-to-infrastructure reading — Axiom R as the One Record engine — is AJ's interpretation of G's work, not yet a theorem. For it to bear doctoral weight it must be formalized, either by G or by a rigorous derivation of the structural equivalence. Today it is grounded scaffolding, honestly named — not a finished bridge.
3The proposed vehicle: PhD by publication
PhD by publication (and its prior-published-work variants) is the route where the dissertation is the portfolio: the gate is peer-reviewed outputs — several, not one — plus a coherence narrative binding them into one contribution, plus an institution willing to examine it. That fits this programme exactly, because the coherence narrative already exists as running code: the spine in §2 is not a chapter to be written, it is a system to be pointed at.
The honest timeline, named as an estimate: peer review commonly takes months to a year-plus per paper, and a portfolio needs several. This is a 2–4 year road. The estate ships in days; journals do not — and no amount of shipping velocity changes a journal's queue. Papers first. Receipts always.
4The receipts so far — live, re-runnable
Everything below is public, live now, and re-runnable offline; nothing asks for trust. These are offered as case studies of the method — falsification conditions that actually execute — not as physics results (see wall 5).
| Receipt | What it shows |
|---|---|
| a₀ research note → | G's parameter-free form a₀ = C_S²·cH₀/(2π) ≈ 1.19×10⁻¹⁰ m/s² against MOND's fitted ≈1.2×10⁻¹⁰ — a residual of −0.67% at H₀ = 73.8 km/s/Mpc, and H₀-dependent by construction (that dependence is its sharpest published kill switch). Independently reconstructed and re-run. |
| SPARC galaxy test → | The full survey, both scatter numbers stated: raw 0.199 dex over 3,389 points from all 175 galaxies, none excluded; 0.144 dex over 2,803 points under the survey's own published velocity-error cut (robust σ 0.143 raw / 0.122 cut). The published ~0.11–0.13 dex band (McGaugh et al. 2016) is a post-cuts number — the honest comparison is cut-to-cut. |
| collider ladder → | Parts I–III vs CODATA: mp/me to 0.010 ppb (tolerance 0.017 ppb), G to 0.69% (tolerance 1.0%), the neutron–proton mass difference to 2.23 ppm (tolerance 5.0 ppm) — all inside G's published tolerances. Not a simulation; no CERN affiliation. |
Both physics receipts run offline, Python standard library only, and exit non-zero if any number drifts outside its stated gate:
# from the lucid-rodeo repo — no network, no dependencies
python3 the420code/proofs/galaxy-rotation/demo.py
python3 the420code/proofs/particle-bench/demo.py
What the collider ladder is — and what it is not
Not a simulation, and no affiliation with CERN or any accelerator programme — nothing there models a beam, a detector, or an event. The "measured" side is CODATA: the distilled output of decades of real precision experiments, confronted with closed forms that never saw them.
And the register, held throughout: reproducing the arithmetic is internal consistency, not proof that nature agrees. Whether the closed forms are physics or numerology that happens to land is a question for specialists — and it is not the question this programme stakes its doctorates on.
5 · The walls, plainly
A proposal note that hides its walls is marketing. These are the walls, in the order a sceptic would raise them:
- G has not agreed to this. The two-doctorate framing is AJ's proposal. G's published corpus is discussed here because it is public (DOI above, CC BY 4.0); G's pursuit of any doctorate is not — it is proposed, not agreed, and this page commits G to nothing. Until G's explicit word says otherwise, G's lane is an invitation, not a plan.
- No institution yet. Nobody is enrolled anywhere. There is no supervisor, no programme, no examiner. Everything on this page precedes all of that.
- Nothing submitted yet. Zero peer-reviewed outputs exist. Every receipt above is live and re-runnable — and none of it is yet citable. The road runs through peer review, not around it.
- Every submission is gated on consent. Nothing touching G's corpus moves — not one submission — without G's explicit consent. That is covenant, not courtesy, and it is a hard gate, not a formality.
- The physics claims are case studies of the method. The doctoral claims are executable falsifiability and self-verifying systems. The physics receipts demonstrate that the method runs on a real, published corpus; they are not claims that must survive physics referees, and this page does not ask them to. Consistency is not proof of physics — here, or anywhere on this estate.
If any of these walls proves taller than stated, this page will say so. That is the whole posture.