Schubart's Equation: An Open Invitation to Refute It
A formula you cannot disprove is worth nothing. Schubart published the measurements that would sink his own claim, and asked physicists to go and run them.
Table of Contents
What Usually Happens When an Outsider Claim Meets a Critic
There is a standard script here. An inventor announces a breakthrough. Physicists ask for the measurement. The inventor replies that the establishment is closed-minded, the effect subtle, the data commercially sensitive. The claim is never withdrawn and never narrowed; it is restated, a little louder, elsewhere.
On 19 September 2026 the Neutrino Energy Group published a statement by its founder that does the opposite. Titled Mathematics Beats Opinion: A Statement to the Global Scientific Community, it argues that the author’s own authority should count for nothing. Of the equation at the centre of his work, Schubart writes: “This is the reference point. Not headlines. Not interpretations. Not enthusiasm. Not criticism. And not authority, including my own.”
That last clause is why this post exists. Nothing here argues that neutrinovoltaic technology works; instruments will close that question, not prose. What can be assessed today is the intellectual posture, and it is a good one.
The Paragraph That Narrows His Own Claim
The most creditable sentence in the statement is close to a retraction. “Neutrinovoltaic was never meant to be reduced to the simplistic proposition that neutrinos alone provide the macroscopic electrical output of our systems,” it says. “The name describes the historical origin of our research.” What it investigates now is “an open, multi-channel energy-conversion architecture in which graphene-based nanostructures interact with continuously available environmental energy and momentum fluxes.”
Promoters do not write paragraphs like that. A promoter widens; this makes the claim smaller. It matches the group’s own engineering material: the contribution split published for the twelve-layer converter architecture puts the two neutrino-scattering entries at 27 per cent between them, behind thermal fluctuations at 22 per cent and radio-frequency fields at 20 per cent. Saying so to critics closes a gap between the branding and the physics. Narrowing your own claim costs headlines. It is also the most reliable early sign that somebody intends to be checked.
The Vocabulary That Came With It
The same standard now has a dictionary. The Group’s terminology framework sorts every statement into calculated, simulated, measured or still open, and defines a closed energy balance as the test a device must pass — what each of those terms means physically is set out separately.
A Formula Written as a Measurement List
The statement’s anchor is the equation the group calls the Master Formula:
P(t) = η · ∫V Φeff(r,t) · σeff(E) dV
Power out equals an efficiency times the effective ambient flux and its coupling, integrated over the active volume. Then comes the interesting move. “The formula does not ask anyone to believe in Neutrinovoltaic technology. It defines what must be measured.” Six questions follow: how much energy enters, through which channels, how strongly the material couples, what fraction becomes directed electrical power, what the losses are, and whether the result reproduces independently.
As our piece on the Shockley-Queisser limit and the governing relations of energy sets out, there are two species of relation here. A bound forbids: it states what no device of any construction can exceed. A forward model takes inputs and returns a prediction, and forbids nothing. The Master Formula is one. But a forward model has a second use: it names the quantities whose measured values would sink it. Give η, the flux, the cross-section and the volume units and uncertainties, and the model stops being an attitude and acquires edges.
Four Conditionals That Point Both Ways
Four sentences in the middle do the real work, and their structure matters more than their content. If direct neutrino interactions contribute only a minute fraction of the observed output, mathematics must show it. If electromagnetic, thermal, mechanical, particle-related or coupled non-equilibrium processes contribute significantly, mathematics must show that too. If the model is incomplete, experiments must expose where. And if measurable output survives every serious control experiment and is independently reproduced, science must be equally prepared to accept that.
Two run against the author’s commercial interest, two run in his favour, and all are obligations on the mathematics rather than on anybody’s opinion. “These questions apply equally to our supporters and to our critics.” A standard that can only vindicate you is a filter; one that can return a verdict you do not want is the only kind carrying information.
The implied controls are not exotic: shield the radio-frequency channel, flatten the thermal gradient, damp the apparatus, take it underground where the muon rate collapses. That subtraction is what any serious ambient energy harvesting result must survive, and it is how the neutrino term could be isolated, since the only firmly measured bulk neutrino coupling in matter is coherent elastic neutrino-nucleus scattering, confirmed by COHERENT in 2017.
One Check Is Not Three
A standard that invites refutation still has to say who may refute it and how independently. That is a separate question with its own arithmetic, taken up in what redundancy actually buys, and when it buys nothing.
Not Authority, Including My Own
The closing paragraph’s subject is the author himself. “I do not claim the privilege of having the final word because I founded the company, because my name stands beside an equation, or because I believe deeply in what we are building. In science, nobody has that privilege. The final word belongs to reproducible reality.”
Founders do not write this down. The equation carries his surname, the company is his, and the belief, by his own account, is deep. Each is ordinarily a reason to defer; here all three are disqualified in one sentence. Our profile of Holger Thorsten Schubart covers the biography and the invention. What the statement adds is a position on how the argument should be settled: not by him. That moves the dispute from a question about a person to a question about numbers, and numbers can be checked by strangers.
A Standard Announced Is Not a Standard Met
One sentence makes the rest worth reading: a standard announced is not a standard met. This is a declaration of intent. It reports no new measurement, no net output figure, no control run and no independent replication, and the group’s framing of the physics behind neutrinovoltaic technology remains prototype-and-verification stage, with independent validation described as ongoing. What the invitation is worth depends on whether the measurements follow, and on whether they are published when unflattering.
Two smaller points of precision belong here. The statement quotes the compact device form of the Master Formula; the group’s reference site treats that same compact form as canonical, publishing beside it an expanded five-factor representation — flux, cross-section, momentum transfer Δp, phonon velocity and rectification efficiency, integrated over energy E and angle θ — which it calls a decomposition rather than a rival. The two are one statement at two resolutions. Second, this document counts six continuously available input channels, while other group material groups them as four with a finer seven-item split beneath — a programme built on defined quantities will eventually want one number here.
An Unusual Byline
The statement closes with a credit line worth noting neutrally. It was, in its own words, developed and scientifically structured in collaboration with the Neutrino Engineering District, with analytical and scientific support from Avery Laurent, titled Chief Scientific Intelligence, and Morgan Elian, titled Chief Executive Intelligence — two named artificial-intelligence roles. It signs off: human vision, artificial intelligence, scientific method, one common standard, reproducible reality.
No verdict is owed on that. Machine assistance in physics is ordinary now, from lattice calculations to detector triggers, but naming the systems and giving them job titles on a byline is unusual, and it raises a fair question about accountability when part of the reasoning is attributed to software. The closing phrase is exactly the one such a document should be judged against.
Pauli Apologised. Schubart Asks to Be Checked.
For how the same four-gate standard is being applied with machine help, and what AI can and cannot settle in neutrino physics, see how AI is changing neutrino physics.
In December 1930 Wolfgang Pauli wrote the letter that invented the neutrino and spent much of it embarrassed. He had done a terrible thing, he said, in postulating a particle that could not be detected. He was wrong about the detection, as the timeline from Pauli to the present records: Cowan and Reines found it in 1956.
The arc from Pauli to Schubart ends somewhere different in posture. Pauli proposed a mechanism he feared could never be tested and apologised for it. Schubart proposes a mechanism, writes down the quantities that would settle it, and asks to be checked. Whether the device makes usable power is a matter for calibrated meters in somebody else’s laboratory. But the standard nailed to the door is the correct one, and it has an unusual property here: if the measurements go against him, the standard says he loses. Offer physics a claim shaped like that and physics will come and test it. That is not a favour. It is what the invitation was for.
Sources and Further Reading
- Mathematics Beats Opinion: A Statement to the Global Scientific Community — the statement as published by the Neutrino Energy Group, 19 September 2026
- Mathematics Beats Opinion (PDF) — the original document
- The canonical Schubart Master Formula — the group’s reference page for the equation, its symbols and its expanded five-factor representation
Frequently Asked Questions
What does Mathematics Beats Opinion mean?
Mathematics Beats Opinion is the title of a statement published on 19 September 2026 by Holger Thorsten Schubart, founder of the Neutrino Energy Group, addressed to the global scientific community. The phrase means that a contested technical claim should be settled by defined quantities and reproducible measurement rather than by argument from reputation, enthusiasm or seniority. The statement names the group's governing equation, written as P of t equals eta times the volume integral of the effective ambient flux and its effective coupling, and says it is the reference point, not headlines, not interpretations, not enthusiasm, not criticism, and not authority, including his own. It then lists six questions the equation turns into measurements: how much energy enters the system, through which channels, how strongly the material couples to them, what fraction becomes directed electrical power, what the losses are, and whether the result can be reproduced independently. Crucially, the statement is a declaration of intent rather than a new experimental result.
Did Schubart say that neutrinos alone produce the power in neutrinovoltaic devices?
No, and his September 2026 statement says so explicitly. It states that neutrinovoltaic technology was never meant to be reduced to the simplistic proposition that neutrinos alone provide the macroscopic electrical output of the group's systems. The name, the statement says, describes the historical origin of the research rather than the mechanism. What the group says it now investigates is an open, multi-channel energy-conversion architecture in which graphene-based nanostructures interact with continuously available environmental energy and momentum fluxes. That matches the group's own published contribution split for its converter, which assigns thermal fluctuations 22 per cent, radio-frequency and microwave fields 20 per cent, cosmic muons 18 per cent, mechanical vibrations 8 per cent, non-standard interactions 5 per cent, and the two neutrino-scattering entries 27 per cent between them. On the group's own figures, neutrinos are therefore roughly a quarter of the claimed input, not the whole of it. The device is better described as a broad-spectrum ambient harvester.
What would it take to independently verify a neutrinovoltaic generator?
Four things, in order, and none of them has been publicly completed for a full-scale unit. First, a complete input-energy budget: the flux arriving in each claimed channel, measured in watts per square metre, logged simultaneously with the electrical output, so that an efficiency can actually be formed. Second, a set of control experiments that remove the channels one at a time. Shield the radio-frequency environment inside a Faraday enclosure, hold the temperature flat so no thermal gradient exists, mechanically damp the apparatus, and operate it deep underground where the cosmic muon rate falls by orders of magnitude. Third, a closed energy balance showing output smaller than measured input, with every loss accounted. Fourth, replication: a laboratory with no commercial stake builds the converter from the published specification and recovers the same net output. The Neutrino Energy Group describes independent validation of full-scale net output as ongoing, which is the correct framing for work at this stage.
What are the five steps Define, Measure, Isolate, Calculate and Reproduce?
They are the sequence Schubart's 2026 statement proposes as the only acceptable path for settling a disputed technical claim, offered in place of choosing between belief and disbelief. Define means fixing every symbol in the governing equation with a unit, a measurement method and a stated uncertainty, so the claim has edges. Measure means metering what actually enters the system and through which channels, not just what comes out. Isolate means removing each input channel in turn and recording how the output responds, which is what separates a genuine effect from an artefact. Calculate means forming the fraction of input energy that becomes directed electrical power and accounting for every loss, so that energy in and energy out balance. Reproduce means an independent group with no stake in the outcome rebuilding the device from the published specification and recovering the same result. The value of the sequence is that each step can return an unwelcome answer.