I've followed your public work for years — Discovery, YouTube, short interviews — on and off, not comprehensively. When you talk about why falsification matters — not peer applause, not elegance, not precedent — that emphasis is close to the methodology I've tried to build into this framework, which is part of why I'm writing to you in particular.
So here's a direct proposition: I have 43 numbered predictions with explicit falsification conditions. One is publicly ❌ falsified; one is ⚠️ partial — the ordering confirmed, the specific residual claim not detected. Both are on the site, written before the tests. I'd like you to pick another one to stress-test.
The opening I'd use with you
Proposed opening — Option B
"One of my predictions is publicly stamped ❌ falsified, one is ⚠️ partial — both on my peer-review page, both written before the tests. I'd like to talk about a framework where being wrong publicly is part of the methodology — not a weakness."
From what I've seen of your public work, intellectual honesty seems to weigh more than polish — that is the part I'd lean on. The ❌ stamps on the peer-review page are not embarrassments; they're the evidence that the falsification conditions were real and were met.
One stamped, one partial — both written before the results
RF 1.3 MHz bird magnetoreception
The framework predicted RF 1.3 MHz would disrupt cryptochrome-based navigation. Schwarze et al. 2016 showed the disruption occurs at fundamentally different frequencies. Falsification condition met. Stamped: date + verbatim quote from paper.
Compton ordering + magic-number residuals
Compton ordering confirmed (R²=0.80, four known Z-inversion anomalies reproduced). But magic-number residuals: Mann-Whitney p=0.29, independent Welch t=+0.51, p=0.63. No significant signal. The specific residual claim failed; core ordering holds.
One ❌, one ⚠️. Falsification conditions were written before the tests, not after.
The prediction I first meant to put in front of you — now on hold
Resolution-dependent deviations at LHC energies
Layer L1 · Open · test suspended while the derivation is under review
The original claim: particle collisions at LHC energies would show non-Gaussian patterns in position measurements beyond Heisenberg uncertainty alone. Its derivation assumed a fixed Planck lattice. My current working choice is a random substrate of fixed density, with no minimum length, and whether that produces any such deviation is open. Even on the lattice reading the effect scales as (E/E_Planck)², about 10⁻³⁰ at 13 TeV if its coefficient is of order one: far below any collider. At that size an LHC null result could not test it either way, so I have suspended the test rather than keep presenting it as live.
If you see a version of this that a particle physicist could actually test, that is one of the conversations I'd most like to have.
Full prediction list with falsification conditions →Second candidate
Schumann × L-chirality. Bacterial protein synthesis in a 7.83 Hz AC field vs control, chiral HPLC analysis. Honest epistemic status: Category C — lens-inspired hypothesis, not derived from core postulates. Explicitly marked as such on the site. A PhD experiment in an afternoon if someone wants to run it.
What I'm not claiming
Not a peer-reviewed paper. It's a preprint on Zenodo (v0.1, May 2026). It predates several corrections made on this site since; the site text is the current version. The manuscript is there for scrutiny, not as a credential.
Not "I've discovered physics." The framing on the About page is explicit: "personal framework, not established physics."
Not complete formalism. Four explicit limits are listed at /about/#honest-limits — including "incomplete mathematical formalism" and "analogy risk."
Not consciousness-causes-collapse. The filter hypothesis in the framework follows Bergson → McFadden, explicitly not von Neumann–Wigner.
"What does this add over QFT?" — I expect you to ask this
QFT describes what happens operationally and does it brilliantly. The substrate model proposes an explanation for why specific values occur — c, three fermion generations. QFT takes these as free parameters; I propose a framework where they're in principle derivable.
Prediction #26 is the concrete test: proton mass from grid parameters alone (in my current working choice the substrate's density is itself a free parameter of Planck order). If it works, there's added value. If it doesn't, the model fails at that layer. Occam is respected — no more parameters than QFT, different ontology.
What I'm asking for
"Which prediction do you think fails fastest?
Let's start there."
One prediction. One falsification condition already written. One honest conversation about what the data says. If it gets a ❌ — great, that's the process working. If it doesn't fail — maybe prediction #2 is worth looking at.
For the record
- Author Marald Bes — spectrumofeverything.com/about/
- ORCID 0009-0009-7697-2811
- Zenodo DOI 10.5281/zenodo.20043846 (preprint v0.1 — the site text supersedes it)
- Peer Review spectrumofeverything.com/peer-review/
Personal framework — not established physics. Disclaimer on /about/.