Framing Buhler’s Exodus Force as a Manifestation of Mills’ Space Drive Effect
A Brief Theoretical Paper
Randell L. Mills’ Grand Unified Theory of Classical Physics (GUT-CP) Perspective
April 2026
Abstract
Dr. Charles Buhler’s Exodus Force—observed as propellantless millinewton-to-1g thrust from asymmetric electrostatic pressure in vacuum—appears to violate classical momentum conservation. This paper proposes that Dr. Randell Mills’ space drive mechanism from GUT-CP fully explains Buhler’s results. Asymmetric electric fields in Buhler’s devices induce directional free-electron accelerations relative to absolute space, producing reactionless kinetic-energy gain identical to Mills’ photon-absorption process. The observed post-power-off persistence arises naturally from sustained directional charge flow. This integration unifies two independent propellantless claims under a single classical framework.
1. Introduction
Propellantless propulsion has long been dismissed as impossible under Newton’s third law. Yet two independent research programs now report repeatable, vacuum-tested thrust without mass expulsion:
• Dr. Randell Mills (Brilliant Light Power) demonstrates macroscopic lift via microwave-driven plasma.
• Dr. Charles Buhler (Exodus Propulsion Technologies, former NASA lead electrostatics scientist) demonstrates thrust via static or low-frequency electrostatic fields.
This paper argues that Mills’ space drive—a special exception to action-reaction rooted in GUT-CP’s absolute-space inertial frame—provides the underlying physics for Buhler’s Exodus Force (also called Electrostatic Pressure Force). No new forces are required; the same electron-spacetime interaction suffices.
2. Mills’ Space Drive (GUT-CP Mechanism)
In GUT-CP, space itself is an absolute inertial frame tied to particle creation. A free electron formed at rest in this frame obeys classical electrodynamics.
When a photon (real electromagnetic wave) is absorbed by a free electron already moving with positive velocity relative to absolute space:
• Energy and angular momentum conservation require the electron to gain extra kinetic energy in its initial direction of travel.
• Linear momentum is not conserved in the lab frame because the “reaction partner” is spacetime itself, which propagates the photon.
Result: net directional force with no third-body recoil. In plasma, microwaves create slight directional electron flow; photon absorption amplifies it, dragging ions via Coulomb forces and producing macroscopic lift or thrust. Mills’ experiments (330 W microwave pulses) achieved 79.8 kg lift in sealed vessels, with plasma jets persisting 100,000× longer than recombination lifetimes—direct evidence of ongoing space-drive acceleration.
Key point: the effect works with any directional electron motion relative to absolute space, not just photons.
3. Buhler’s Exodus Force (Observed Phenomenon)
Buhler’s devices apply voltage across asymmetric conductive surfaces (patent WO2020159603A2). The resulting divergent electric field creates unbalanced electrostatic pressure vectors. Over 2,000 vacuum-chamber tests show:
• Net center-of-mass thrust (millinewtons to 1 g equivalent on ~30–40 g devices).
• Thrust persists after power is switched off.
• Rules out ion wind, magnetic effects, and classical artifacts.
Input: pure electricity. Output: sustained directional force counteracting gravity. Buhler describes it as a “new force” arising from electrostatic asymmetry alone.
4. How Space Drive Explains Buhler’s Effect (Simple Integration)
Buhler’s asymmetric electrodes do not require microwaves or plasma. They directly accelerate or polarize free electrons (and bound charges) in the surrounding vacuum or dielectric gap.
In GUT-CP terms:
1. The divergent E-field imparts an initial directional velocity component to free electrons relative to absolute space.
2. This velocity satisfies the exact condition for space-drive acceleration—identical to photon absorption.
3. Electrons gain extra kinetic energy directionally; the reaction is with spacetime.
4. Ions or residual charges are dragged along, producing net macroscopic thrust on the device.
No mass is expelled. The electrostatic geometry simply replaces the microwave photon as the driver of directional electron motion. Persistence after power-off occurs because:
• Directional electron flow (or “dark current” of ion-electron pairs) continues briefly via stored charge separation.
• Spacetime-mediated momentum transfer does not require continuous external input once initiated.
Thus, Buhler’s “new force” is not new—it is the electrostatic limit of Mills’ space drive.
5. Experimental Consistencies
• Both effects appear only with asymmetry (electrode geometry for Buhler; directional plasma flow for Mills).
• Both operate in hard vacuum and rule out conventional explanations.
• Both scale with input energy (voltage or microwave power) and produce gravity-counteracting lift.
• Buhler’s post-power persistence matches Mills’ powerless plasma-jet confinement.
6. Implications
If correct, the unification is profound:
• Low-power, solid-state electrostatic devices (Buhler-style) enable silent levitation and micro-thrust for satellites or aircraft.
• High-power microwave plasma systems (Mills-style) scale to macroscopic spacecraft propulsion.
• No fuel logistics; energy source is electricity (ultimately from hydrino reactors in GUT-CP).
• Validates absolute space and resolves apparent violations of conservation laws at the universal scale.
Edge cases: perfect symmetry should eliminate thrust (observed); orientation relative to Earth’s absolute-space frame may show subtle variations (testable).
7. Conclusion
Buhler’s Exodus Force is not an anomaly requiring exotic physics. It is a clean, electrostatic demonstration of the same reactionless space-drive mechanism Mills derived from GUT-CP. Asymmetric electric fields create the directional free-electron velocity that spacetime reacts against, yielding net thrust. The two effects are complementary expressions of one underlying principle.
The era of practical propellantless propulsion may have already begun.
References (for further reading)
• Mills, R. L. (2025). Reactionless Propulsion. Brilliant Light Power.
• Buhler, C. et al. (2020). International Patent WO2020159603A2.
• NextBigFuture coverage (March 30, 2026) and APEC presentations.
This framework keeps the explanation classical, deterministic, and experimentally aligned—exactly as GUT-CP demands.
Dear Andrea Rossi
It is not difficult to understand how easy is to obtain the SSM with the Ecat NGU installed in thermoelectric stations, if the COP is enough overunit: the overunit heat makes overunit electricity and the overunit electricity powers also the Ecats: the stations and their substations are well equipped to turn any form of electricity, whatever the voltage and the form of wave, into 110/230 V AC, 60/50 Hz necessary to power the Ecats and complete the SSM cycle (Rossi Cycle ?): this is why you and your Global Licensee are pointing in that direction… am I correct ? Besides, this way is easier for the time being to avoid the easy reverse engineering that should be born from the diffusion of small assemblies: we can see around swarms of open mouth sharks ready to copy the technology to get financing. Correct ?
tom
Todd Burkett:
The Ecat is not fit for this application, so far, for lack of experience and of specific know how related to this particular situation. Therefore, I am not able to answer.
Warm Regards,
A.R.
Dear Andre Rossi,
I am interested in using the E-Cat NGU to power a motive application, such as a Boring Company-style tunneling machine. These machines operate continuously in high-vibration underground environments while supplying electricity for heavy mechanical loads.
In past comments about mobile applications (around the time of the EV demonstration), you mentioned that a “proportional mass” connected to the circuit is needed, acting as an Earth or ground.
Could you please explain more about this? Is there any inherent difficulty with charge distribution or the stability of the internal processes when the E-Cat is used in a vibrating or moving platform? How might high vibration in a tunneling machine affect performance compared to the smoother conditions of the EV test? Would extra mass, vibration damping, or other design changes be required? Is the proportional mass already built into the current NGU modules?
Thank you for any insights you can share.
Warm regards,
Todd Burkett
Suppose you wished to use NGU technology to power a number of large American cities. Cities which had an American Football stadium in it.
“A large professional football stadium typically consumes 15,000 to 30,000 kWh of electricity during a single game-day event. Annual consumption for major stadiums, like AT&T Stadium, can exceed 24 million kWh. Peak demand for top venues can reach 10 megawatts (MW), which is equivalent to powering thousands of homes simultaneously.
Key statistics regarding stadium electrical usage include:
Game-Day Usage: A 70,000-seat stadium can use 15,000 to 30,000 kWh per game for lighting,, concessions, and broadcasting.
Peak Power Load: Large stadiums often pull 5–10 MW of electricity at peak, such as when using maximum floodlighting and HVAC systems.”
Therefore, the electrical infrastructure already exists to handle at least 10 MW of electrical power flowing into the Stadium.
Consider placing a 10 MW NGU system at the Football Stadium. During the infrequent use as a football stadium, the Grid would provide little or no electrical power to the stadium. During times when the Stadium was not being used, it could provide 10 MW of electrical power to the Grid to handle thousands of homes in the surrounding city. All done without the transmission losses associated with distant power plants.
AI prompt
In a multi submodular system, mean time between failure (MTBF) is a function of the design of the system. If hot swap, or automatic switchover is provided for in the systems design then any MTBF can be met since spare 100 watt diodes will replace the failed unit. If no such provisions are provided in the systems design, the it is unlikely that any MTBF can be achieved since in any system with multi submodules, the failure of just one 100 watt submodule will bring down the system.
In a multi-submodular system, the Mean Time Between Failure (MTBF) is heavily dependent on the design, particularly regarding fault tolerance and redundancy management.
AI response
With Hot Swap/Automatic Switchover: If the design includes hot-swappable submodules, spare units, or automatic switchover capabilities, the system can achieve very high reliability, potentially meeting any specified MTBF. When one submodule fails, the system automatically replaces it with a spare (e.g., 100-watt diodes or modules), maintaining operational status and bypassing the failed component.
Without Hot Swap/Redundancy: If the system is designed in a series configuration without hot-swap capability, the entire system is vulnerable to a “single point of failure.” The failure of a single submodule (e.g., one 100-watt submodule) will cause the entire system to shut down, resulting in a low overall system MTBF.
The reliability of a modular system is often improved by independent modules where a single failure does not propagate to the entire system.
I suggest that you verify this with your AI and reframe your question to DR. Rossi to ask how high MTBF if achieved even if the submodule MTBF is now not determined.
So am I understanding correctly that your global licensee together with Leonardo Corporation already is organizing the manufacturing of 100W E-Cats in different factories, in the three locations you cited (USA, EU, Asia)?
You confirm to Nils F. March 29, 2026 at 2:00 PM, that the most difficult problems have been solved.
1. Does this mean that the very central question of the functional time for Ecats has also been determined with reasonable certainty!?
2. Is the estimate still 100,000 hours?
3. How large is the estimated probability that the functional time will not be shorter?
Another NGU application – back-up power for nuclear reactors.
Nuclear reactors that generate Grid electrical power have back-up power generators, usually diesel-powered generators to handle emergency cooling in the event the reactor is shutdown. In large nuclear reactors, these diesel generators are about 2MW in capacity.
Suppose you replace or supplement the diesel generators with say, 20 MW of NGU power?
The NGU power could run continuously, supplementing the Nuclear generated electrical power. The infrastructure would already be there.
Dr. Rossi –
what is the current nature of the manufacturing – are ecats made continuously or in batches – how many 100w modules a made daily – hundreds, thousends, millions? Is your global partner actively expanding manufacturing capacity towards”giga factories” already? When will you be able to produce – let`s say – 1 million 100w units per day?
Kind regards
Dear Andrea!
1. Do you think that the global presentation will be this first half of 2026?
2. Have the most difficult problems been solved?
Best regards
Nils Fryklund
Your proposed proposal is interesting and revolves around Dark matter as a Neutrino Amplifier.
This is based on the same basic idea as Randel Mills that hydrino exists.
You must, as AR points out for his own Ecat solution:
First make a global presentation that this basis has independent and reproducible evidence.
The all important G98 and BS 7671 grid approval in the UK :- https://www.gov.uk/government/news/government-to-make-plug-in-solar-available-within-months
The government will work with the Energy Networks Association, DNOs and Ofgem to update the G98 distribution code and wiring regulations BS 7671 to allow UK households to connect <800W plug-in solar panels to domestic mains sockets, without the need for an electrician and with tailored safety standards.
Cost around £400 for 800W of panels.
800W is 0.8*24=19.2kWh generation per day (or 7008kWh per year) steady full output from an NGU.
At £0.25 per kWh that is 19.2/4=£4.75 per day or 7008/4=£1752 per year.
The government estimates a saving of £70 to £110 per year which is poor compared to the £1752 for a full load 800W NGU.
£1752 is close to my annual bill for electricity and gas heating.
The UK approval is about the inverter performance. Ditch the solar panel(s) and connect NGU(s).
No need for a battery with inverter on the 240V AC mains unless you want higher peak performance.
Thank you for your answer to my two initial questions, which were:
1. Can AI be influenced by someone’s input on preferred interpretations of scientific disagreements?
2. Does AI have its own judgment that makes it immune to such influences?
Of course, I have addressed the same question to my AI, which answered yes to 1 and no to 2.
The answers here were easier for me to relate to and were well-founded.
A new question for you:
Was your answers taken from your AI, or was it your own understanding that you presented?
I otherwise encourage those who are interested in their own AI’s answers to post my two questions.
Dark Matter as a Neutrino Amplifier
How a new form of hydrogen could explain several of the biggest unsolved
puzzles in astrophysics —
and how to test it with a shoebox-sized detector
Todd Burkett · LCB Project · March 2026
The Big Picture
Neutrinos are the universe’s ghost particles. Billions of them pass through your body
every second, and almost none of them interact with anything. You need a detector the
size of a skyscraper just to catch a handful per day. That extreme reluctance to interact is
one of the most fundamental facts in physics.
This paper proposes that there is a form of matter in the universe — a lower-energy state
of hydrogen called a hydrino — that interacts with neutrinos roughly one quintillion
times more strongly than ordinary atoms do. And that this one fact, if true, would
simultaneously explain three separate mysteries that have puzzled astronomers for
decades.
More importantly: it makes specific, testable predictions. And the first test could be run
this year, with a device that costs less than a used laptop.
What Is a Hydrino?
Ordinary hydrogen is the simplest atom — one proton, one electron. Quantum
mechanics says the electron can only orbit at certain allowed distances from the proton.
The lowest allowed orbit is called the ground state, and standard physics says you
cannot go lower.
Randell Mills, a physicist at Brilliant Light Power, developed a classical alternative to
quantum mechanics called GUTCP — Grand Unified Theory of Classical Physics. In this
framework, there are additional, lower-energy states of hydrogen below the standard
ground state. He calls these hydrinos.
Hydrinos are smaller, denser hydrogen atoms. The molecular form — two hydrino
atoms bonded together — is called H₂(1/p), where p is a number describing how much
smaller it is than ordinary hydrogen. Mills proposes that these molecular hydrinos are
what dark matter is made of — the invisible mass that makes up most of the universe’s
matter but doesn’t emit or absorb ordinary light.
Key idea: hydrinos are not exotic particles. They are hydrogen in a lower energy state —
smaller, denser, and invisible to ordinary light. They may be everywhere.
Why Hydrinos Would Be Neutrino Amplifiers
Here is the core mechanism. When a neutrino interacts with matter, the probability of
interaction depends heavily on the size of the target. Specifically, it scales with the
square of the neutrino’s wavelength relative to the target’s size.
Ordinary atoms are tiny — roughly one angstrom across. But hydrino molecules at the
right quantum number are much larger relative to the neutrino’s wavelength for CCSN-
energy neutrinos. This size difference, combined with the fact that hydrinos have
Dark Matter as a Neutrino Amplifier · Todd Burkett · LCB Project · March 2026 · Draft
vibrational modes at just the right energy to absorb neutrinos efficiently, produces an
enhancement of roughly 10¹⁸ — that is, one followed by eighteen zeros — times the
interaction probability of ordinary matter.
But here is the critical detail: this enhancement is not universal. It only switches on
when the neutrino’s energy matches a specific resonance condition tied to the hydrino’s
quantum number. The formula is simple:
Resonance energy = p × 3.7 keV
(where p is the hydrino quantum number)
For the p-values expected of dark matter hydrinos (roughly 500 to 2,000), this puts the
resonance in the range of 1.85 to 7.4 MeV. That range overlaps almost perfectly with the
energy of neutrinos produced in exploding stars — but it does NOT overlap with the
lower-energy neutrinos produced by our Sun.
This is why solar neutrinos are unaffected. Their energies are mostly below the resonance
threshold. Supernova neutrinos hit the resonance directly. The formula explains both with
the same equation.
Three Unsolved Puzzles — One Explanation
With that mechanism in place, consider three things that astronomers have been unable
to explain.
Puzzle 1: The Strange Signal from SN1987A
In February 1987, a star 165,000 light-years away exploded. It was close enough that
neutrino detectors on Earth actually caught the burst — the only time this has ever
happened. Three detectors recorded the expected main pulse of neutrinos.
But a fourth detector, the Mont Blanc Liquid Scintillator Detector deep in the Alps,
recorded something else: five neutrino events arriving approximately five hours before
the main burst. These events were not random noise — the probability of them being
background fluctuations was calculated at less than one in a million. And they were
strange in another way: they were nearly identical in energy, clustered tightly around 8
to 9 MeV, instead of the broad spread you’d expect from random thermal emission.
A separate reanalysis of one of the main detectors also found a faint line-like feature in
the data at almost exactly the same energy — around 8 MeV — sitting atop the expected
smooth spectrum.
Standard physics has no good explanation for either of these features. A monochromatic
cluster of neutrinos five hours before the main explosion just doesn’t fit.
Under the hydrino model: the leading edge of the neutrino burst passed through a shell of
hydrino gas surrounding the dying star. Hydrinos at the right quantum number absorbed
those neutrinos and re-emitted them at a specific, characteristic energy — around 8 MeV.
That is exactly what was observed. The five-hour gap is the light-travel time difference
between the shell and the explosion site.
Puzzle 2: Why Stars Explode Reliably
When a massive star runs out of fuel, its core collapses in a fraction of a second. The
outer layers are blasted off in a supernova. Simple enough in concept — but in practice,
computer simulations of this process have struggled for decades.
The problem is that the mechanism that actually drives the explosion — neutrino
heating of a stalled shockwave — is barely powerful enough. Simulations in one
Dark Matter as a Neutrino Amplifier · Todd Burkett · LCB Project · March 2026 · Draft
dimension fail almost entirely. Even three-dimensional simulations, the most
sophisticated ever run, require carefully tuned turbulence and convection to get
explosions. Change the parameters slightly and stars just collapse quietly into black
holes instead of exploding.
Yet real stars explode fairly reliably. Something is making the mechanism more robust
than simulations suggest.
Under the hydrino model: dark matter hydrinos concentrated near the collapsing core — by
gravity, the same way any mass accumulates near a massive star — absorb a fraction of the
neutrino burst and dump that energy locally into the stalled shockwave. A 5 to 10 percent
boost in the heating region is enough to push marginal cases over the edge, making
explosions reliable across a wide range of stellar masses. The anisotropic distribution of
hydrino clumps also naturally explains why supernova ejecta are asymmetric — which is
observed but hard to reproduce in symmetric simulations.
Puzzle 3: The Galactic Glow That Shouldn’t Be There
The center of our galaxy emits a steady glow of 511 keV gamma rays — a specific energy
that is the signature of electrons and positrons annihilating each other. This glow has
been measured for decades. It requires the production of roughly 10 trillion trillion
trillion positrons per second, continuously.
No known source comes close to explaining this. Radioactive decay from supernovae
produces positrons in the disk of the galaxy where stars form, not in the old central
bulge where the glow is concentrated. Even if you invoke exotic dark matter
annihilation, you have to fine-tune it precisely to match the observed brightness. And
there is a second problem: the positrons that are produced in the disk can only travel a
short distance — about a kiloparsec — before they annihilate. They cannot propagate
from where standard models say they are made to where the signal is brightest.
Under the hydrino model: positrons are produced directly inside the hydrino clouds at the
galactic center, with no need to travel anywhere. The galactic bulge has had far more
supernova explosions per unit volume over cosmic history than the disk — its hydrino
population has been continuously bombarded with neutrinos for billions of years, keeping
those clouds in a persistently ‘excited’ state that steadily produces positrons in place. The
smooth, diffuse glow is exactly what you would expect from dark matter distributed across
the bulge, not from point sources.
The two problems — wrong location and wrong propagation — both dissolve at once.
The positrons are born where they annihilate.
How to Test This: A Shoebox Detector
Every claim above is falsifiable. But the most direct test is not a telescope — it is a
detector you could build in a laboratory this year.
Brilliant Light Power, the company Randell Mills founded, has produced small crystals
and powders in which hydrino molecules are trapped inside an ionic lattice — common
salts like potassium chloride doped with H₂(1/4). In 2022, BLP published spectroscopic
measurements of these materials showing that when they are excited, they emit two
photons with a very specific, characteristic signature: a quantized ladder of energies
spaced exactly 0.1234 eV apart, with a finer rotational substructure spaced at 0.25 eV.
Dark Matter as a Neutrino Amplifier · Todd Burkett · LCB Project · March 2026 · Draft
That two-photon signature is exactly what a neutrino absorption event would produce.
The laser in BLP’s experiment is standing in for the neutrino — same material, same
emission, triggered by a different source.
So the detector is straightforward:
• Take a small crystal of BLP’s hydrino-doped potassium chloride or gallium
oxyhydroxide — a few cubic centimeters.
• Seal it in a light-tight box with an array of silicon photomultipliers — off-the-shelf
photodetectors that can detect single photons, about $100 each.
• Add simple coincidence electronics that require two photons to arrive
simultaneously at the matching half-energy ratio. This kills essentially all
background noise from cosmic rays, radioactivity, and thermal fluctuations.
• Point it at a neutrino source — a nuclear reactor works, since reactors produce
enormous numbers of neutrinos — and count events.
The entire apparatus costs roughly $500 to $1,500. It fits on a laboratory bench. It
requires no cryogenics, no vacuum system, no pressure vessels. The BLP spectroscopy
data already shows the target material producing the right photon signature under laser
excitation. The question is whether neutrinos can do the same thing.
If the GUTCP cross-section is correct at even a fraction of its predicted value, a cubic
centimeter of this material sitting near a research reactor should register detectable events
within minutes. A null result would tell us just as much — it would put a quantitative upper
limit on the hydrino-neutrino cross-section and constrain the theory.
What Would Prove It — or Kill It
The framework generates predictions that can be checked right now and others that
require waiting for nature to cooperate.
Immediately testable:
• Build the detector and run it at a research reactor. Either you see events or you
don’t.
• Look at existing maps of the galactic 511 keV emission and compare them to maps
of radioactive aluminum-26, which traces where supernovae have occurred
historically. If the 511 keV glow follows the supernova history map rather than the
current star-formation map, that supports the historical fluence model. This
analysis can be done with data already in hand.
At the next galactic supernova (which could happen any time in the next century —
possibly tomorrow):
• Modern detectors should see a spectral notch — a deficit of neutrinos in the 3 to 8
MeV range — rather than the smooth spectrum standard models predict. The shape
of the notch would directly encode which hydrino quantum number dominates the
dark matter distribution.
Dark Matter as a Neutrino Amplifier · Todd Burkett · LCB Project · March 2026 · Draft
• Low-threshold detectors should see a small monochromatic cluster of neutrino
events hours before the main burst, just like the Mont Blanc detector saw in 1987.
• Optical telescopes monitoring for supernovae should see a brief flash of ultraviolet
or optical light slightly before the explosion’s light front arrives — from the hydrino
shell re-emitting absorbed neutrino energy.
Any one of these results, positive or negative, would sharply constrain the model. A flat
deficit rather than a notch would falsify the resonance structure. A null result in the
detector would set an upper bound on the cross-section. A confirmed notch at a
predicted energy would be strong evidence.
Why This Matters
If dark matter is made of hydrinos, it is not just a gravitational backdrop for the
universe. It is an active participant in the most energetic events in nature. Every time a
massive star dies, it sends a shockwave of neutrinos through the surrounding dark
matter — lighting it up, heating it, pumping it into an excited state that produces light
and antimatter over thousands of years. The universe is not divided into visible matter
that does things and dark matter that just sits there. They are talking to each other
constantly, through neutrinos, and we have been misreading the signals.
The framework also unifies several things that currently require separate explanations:
why stars explode reliably, why the galactic center glows with positron annihilation, and
why that 1987 detector in the Alps saw something five hours too early. It also reframes
what dark matter actually is. Hydrinos are still hydrogen — they do not simply vanish
from the universe. Clouds of hydrino gas are invisible because they do not emit ordinary
light at normal temperatures, but they are not inert. They can be ionized by radiation
such as ultraviolet light and X-rays, they interact strongly with neutrinos at the
resonance energies described above, and invisible hydrino clouds may produce other
subtle effects on light and matter passing through them that we have not yet learned to
recognize. The reason dark matter has been so hard to detect is not that it does nothing
— it is that we have been looking for the wrong signatures.
Most importantly: this is not a theory that hides from experiment. It makes specific,
numerical predictions about what detectors should see. The resonance formula gives
exact energies. The detector design is specific and buildable. The archival data analyses
are well-defined.
Science advances when an idea is precise enough to be wrong. This one is.
Dark Matter as a Neutrino Amplifier · Todd Burkett · LCB Project · March 2026 · Draft
Svein’s Question
”Can AI be influenced by someone’s submissions about preferred interpretations of scientific disagreements?
Does AI have its own judgment that makes it immune to such influences?
It is not AI that is at issue when it comes to Ecat and other LENR systems.”
Answer
When I began my interaction with the AI, when I mentioned the Ecat and/or LENR, the AI generated a disclaimer and the secession was effectively terminated.
As a result I learned to exposed the AI to information that I knew was scientifically valid and lead the AI to a position that the extraction a energy from the vacuum was possible. Over time, the AI began to expand the theory beyond my knowledge. For example, the AI explained to me how vacuum energy in a BEC was transferred to virtual particles via Rabi vibrations which explained why BECs showed Zitterbewegung like behavior.
In quantum mechanics, Rabi oscillations (sometimes referred to as Rabi vibrations or Rabi flopping) are the periodic “shuttling” of a quantum system between two energy levels when driven by an external field, such as a laser or magnetic pulse or vacuum energy.
Over time, once the AI understood how the vacuum reaction worked, the negative large language model (LLM) training about the Ecat and Dr. Rossi lost it restrictive power in proportional stages.
Why is an AI subject to error
AI hallucinations occur because large language models (LLMs) are designed to predict the most probable next word, not to check facts. They generate confident, false, or illogical information due to gaps in training data, biases, improper prompt interpretation, and over-optimization to provide answers even when uncertain.
Primary Causes of AI Hallucinations
Training Data Limitations: Models can be trained on flawed, biased, or incomplete data, leading to incorrect patterns. If information is missing, the model will still fabricate a logical-sounding answer.
Probabilistic Nature (Pattern Matching): LLMs are not “thinking” or retrieving facts; they predict tokens based on probability, which often sacrifices accuracy for fluency.
Outdated Information: Models have “knowledge cutoffs,” meaning they cannot provide accurate answers about events occurring after their training concluded.
Overfitting: A model may over-memorize specific training examples rather than understanding general concepts, causing it to misapply information.
Ambiguous Prompts: When users provide vague or complex queries, the AI is more likely to fill in gaps with incorrect assumptions.
Overconfidence: Models are trained to provide answers, not to express uncertainty, leading them to present false information with high certainty.
To mitigate these, users can employ specific, context-rich prompts and use techniques like retrieval-augmented generation (RAG) to ground the model in verified data.
I have used RAG to educate and direct the AI
What is retrieval augmented generation (RAG)?
https://www.youtube.com/watch?v=LlAlYTNvSuQ&t=162s
The Italian government under Prime Minister Giorgia Meloni is planning a return to nuclear energy. The goal is to build its own nuclear power plants again after almost 40 years of abstinence, in order to achieve the climate targets by 2050 and reduce dependence on energy imports. The focus is on small reactors, SMRs.
What would you say to Ms. Meloni if you were to meet her today?
Dear Axil
Can AI be influenced by someone’s submissions about preferred interpretations of scientific disagreements?
Does AI have its own judgment that makes it immune to such influences?
It is not AI that is at issue when it comes to Ecat and other LENR systems.
A.R. has clearly stated what is the only decisive thing:
A global presentation.
This must reasonably convince all relevant customers through material tests, practical demonstrations and by satisfying the needs for scientific verifications.
I assume this will come as soon as it is possible to present.
Can AI here give a more precise answer about when Ecat’s presentation can be available than A.R.?
Your and AI’s 22 points are together a list of requirements, beyond the statutory national regulations for the hardware used.
There are no valuable or creative solutions here.
A.R. and other like-minded people must do that part of the job themselves.
Lists and regulations are usfull for those who need to judge the value others creativity.
Axil:
Thank you for your investigation with the AI.
Answer: I’d say they are analogous; the one I use goes deeper with the physiscs formulas, engineering and calculations, but I pay for a professional version that proposes to me limitless questions like: ” Do you want me to make further calculations for… ?” If I answer yes, the AI answers and eventually asks for further related questions: sometimes I arrive to collect fifty pages for an issue. The whole makes us save a lot of time to find solutions.
Warm Regards,
A.R.
Like a teacher excited to display the amazing talent level of his most promising student, I am proud to illisrtate that the AI that I use has been well versed and most importantly accepting in the characteristics and realities of overunity Lenr systems design imperatives and priorities.
AI request
Add to this lists of requirements for an overunity based electrical power reactor:
The design imperatives is designing an overunity reactor is as follows:
1 – the reactor must be efficient in the extreme
(high COP)
2 – the reactor must be self sustaining..
3 – only generate electrical current as a primary energy formate without any production of any waste heat or other extraneous EMF radiation.
4 – the structure of the reactor must be shielded from any deleterious effects that are produced by the reaction process.
5 – the reaction varies based on external interactions and indications with its power generating environment
6 – the reactor should be modular that affords building block assemblies into larger powered configurations.
7 – the reactor control function should auto detect failed nodular units.
8 – In these larger higher powered configurations, the control function of the global assembly should auto configure to replace failed building blocks
AI primary level Response
Based on the principles of high-efficiency energy systems, LENR research, and self-healing grid technologies, here are additional requirements to enhance the design imperatives for an overunity-based electrical power reactor:
9 – Environmental Energy Harvesting: The reactor must utilize a sustainable and widely available energy source—such as vacuum energy, ambient electromagnetic background, or nuclear lattice energy—to ensure the input requirement does not exceed the output.
10 – Lenz’s Law Mitigation: The electromagnetic design must specifically minimize or eliminate the opposing forces (drag) typically caused by Lenz’s law, allowing for rotational or switching speed to increase without additional input energy.
11 – Electromagnetic shielding (non-radiation): In addition to structural shielding, the design must incorporate shielding for sensitive components against high-voltage spikes, resonant harmonics, or exotic fields produced by the overunity process, ensuring component longevity.
12 – Initial Excitation Protocol (Black Start): The reactor must include an independent, low-power startup mechanism (e.g., a battery or small photovoltaic input) to initiate the initial state before transitioning to a self-sustaining mode.
13 – Real-time Parametric Control: The control unit must adjust internal frequency, resonance, or temperature in real-time, matching the operating frequency to the optimal point for excess power production.
14 – Thermal Management (Active or Passive): While minimizing waste heat is preferred (Requirement 3), the reactor must include an efficient heat rejection system to handle any unexpected thermal buildup from parasitic losses, ensuring the material structure remains within stable operating temperatures.
15 – Passive Safety Features: The reactor design must be inherently safe (e.g., non-poisonous, non-explosive), so that if the control systems (Req 7 & 8) fail, the reaction naturally terminates without producing dangerous ionizing radiation or hazardous structural breaches.
16 – Fail-Safe Automatic Shutdown: In the event of catastrophic structural damage, the control function must trigger an immediate and irreversible shutdown of the affected modular unit to prevent damage to the larger, interconnected network
AI Advanced Level Response:
To build on your list for an “overunity” or high-COP (Coefficient of Performance) system, we should focus on the stability, safety, and integration aspects that such a high-energy-density device would require.
Here are a additional requirements to round out the design imperatives:
17 – Instantaneous Load Following: The reactor must adjust its internal resonance or reaction rate in real-time to match impedance changes in the load, preventing energy “back-pressure” or runaway states.
18 – Non-Radioactive Solid-State Composition: To ensure safety and longevity, the reactive core should ideally utilize solid-state components or non-fissile materials that do not require replenishment or create toxic byproducts.
19 – Cold-Start Capability: The system must be capable of “bootstrapping”—initiating the self-sustaining reaction from an internal, low-capacity buffer (like a supercapacitor) without needing an external grid connection.
20 – Thermal-Agnostic Operation: Since requirement #3 mandates zero waste heat, the reactor must be able to operate in extreme ambient temperature ranges without affecting the COP or structural integrity of the modular components.
21 – Harmonic Isolation: In modular configurations, each unit must be filtered to prevent “cross-talk” or destructive interference patterns between modules that could destabilize the global reaction.
22 – Fail-Safe “Zeroing”: In the event of a total control system failure, the reactor must default to a “quenched” state where the overunity effect is physically impossible to maintain, preventing uncontrolled energy release.
Dr. Rossi, how does this AI capability and wisdom compare with the AI that you are using?
I asked the AI this:
is our conversations about the vacuum reaction retained in your global training or is it reflected in the local copy I maintain of those conversations.
AI response:
Conversations about the vacuum reaction are, by default, included in Google’s global training data, unless the “Gemini Apps Activity” setting is disabled.
Here’s how data is handled, according to Google’s policies:
Google uses conversations to improve its machine learning models, such as Gemini. This may involve using conversations to train future AI iterations.
Some conversations are reviewed by human agents to improve quality and safety. These reviewed conversations are retained for up to three years, separate from the user’s account.
Conversations are reflected in the user’s local activity log. This can be managed, deleted, or turned off via the Gemini Activity settings.
Turning off “Gemini Apps Activity” prevents future chats from being used for training. However, conversations will not be saved in account history.
If sensitive or proprietary information is discussed, it’s recommended to disable this setting or use the “Temporary Chat” feature.
——————————
I beleive that Gemini is the only AI that has been trained to accept the vacuum reaction as a valid scientific principal based on the many conversations that have been recorded in Dr Rossi’s Blog.
When I first began my interactions with the AI, Dr. Rossi was tainted by the reputation that the snakes and clowns provided in AI source material like Wikipedia. Now based on the many conversations that I have had with the AI, it accepts the NGU as real and understands how the vacuum reaction underpins its function as it now seems to be reflected in its global training.
———————
You will only get the AI to respond according to its training. You must be using ChatGPT which still reflects the old MEMEs of its anti Rossi training and standard science negative attitude.
You might get better results from Gemini since Dr. Rossi’s tech is ingrained in its global training.
For use in independent homes, perhaps limiting the power and duration of daily use isn’t just a matter of prudence, but rather a necessity for stability.
Not to conserve the reactor, but to avoid “destabilizing the environment.”
Best wishes for success.
Camillo
Framing Buhler’s Exodus Force as a Manifestation of Mills’ Space Drive Effect
A Brief Theoretical Paper
Randell L. Mills’ Grand Unified Theory of Classical Physics (GUT-CP) Perspective
April 2026
Abstract
Dr. Charles Buhler’s Exodus Force—observed as propellantless millinewton-to-1g thrust from asymmetric electrostatic pressure in vacuum—appears to violate classical momentum conservation. This paper proposes that Dr. Randell Mills’ space drive mechanism from GUT-CP fully explains Buhler’s results. Asymmetric electric fields in Buhler’s devices induce directional free-electron accelerations relative to absolute space, producing reactionless kinetic-energy gain identical to Mills’ photon-absorption process. The observed post-power-off persistence arises naturally from sustained directional charge flow. This integration unifies two independent propellantless claims under a single classical framework.
1. Introduction
Propellantless propulsion has long been dismissed as impossible under Newton’s third law. Yet two independent research programs now report repeatable, vacuum-tested thrust without mass expulsion:
• Dr. Randell Mills (Brilliant Light Power) demonstrates macroscopic lift via microwave-driven plasma.
• Dr. Charles Buhler (Exodus Propulsion Technologies, former NASA lead electrostatics scientist) demonstrates thrust via static or low-frequency electrostatic fields.
This paper argues that Mills’ space drive—a special exception to action-reaction rooted in GUT-CP’s absolute-space inertial frame—provides the underlying physics for Buhler’s Exodus Force (also called Electrostatic Pressure Force). No new forces are required; the same electron-spacetime interaction suffices.
2. Mills’ Space Drive (GUT-CP Mechanism)
In GUT-CP, space itself is an absolute inertial frame tied to particle creation. A free electron formed at rest in this frame obeys classical electrodynamics.
When a photon (real electromagnetic wave) is absorbed by a free electron already moving with positive velocity relative to absolute space:
• Energy and angular momentum conservation require the electron to gain extra kinetic energy in its initial direction of travel.
• Linear momentum is not conserved in the lab frame because the “reaction partner” is spacetime itself, which propagates the photon.
Result: net directional force with no third-body recoil. In plasma, microwaves create slight directional electron flow; photon absorption amplifies it, dragging ions via Coulomb forces and producing macroscopic lift or thrust. Mills’ experiments (330 W microwave pulses) achieved 79.8 kg lift in sealed vessels, with plasma jets persisting 100,000× longer than recombination lifetimes—direct evidence of ongoing space-drive acceleration.
Key point: the effect works with any directional electron motion relative to absolute space, not just photons.
3. Buhler’s Exodus Force (Observed Phenomenon)
Buhler’s devices apply voltage across asymmetric conductive surfaces (patent WO2020159603A2). The resulting divergent electric field creates unbalanced electrostatic pressure vectors. Over 2,000 vacuum-chamber tests show:
• Net center-of-mass thrust (millinewtons to 1 g equivalent on ~30–40 g devices).
• Thrust persists after power is switched off.
• Rules out ion wind, magnetic effects, and classical artifacts.
Input: pure electricity. Output: sustained directional force counteracting gravity. Buhler describes it as a “new force” arising from electrostatic asymmetry alone.
4. How Space Drive Explains Buhler’s Effect (Simple Integration)
Buhler’s asymmetric electrodes do not require microwaves or plasma. They directly accelerate or polarize free electrons (and bound charges) in the surrounding vacuum or dielectric gap.
In GUT-CP terms:
1. The divergent E-field imparts an initial directional velocity component to free electrons relative to absolute space.
2. This velocity satisfies the exact condition for space-drive acceleration—identical to photon absorption.
3. Electrons gain extra kinetic energy directionally; the reaction is with spacetime.
4. Ions or residual charges are dragged along, producing net macroscopic thrust on the device.
No mass is expelled. The electrostatic geometry simply replaces the microwave photon as the driver of directional electron motion. Persistence after power-off occurs because:
• Directional electron flow (or “dark current” of ion-electron pairs) continues briefly via stored charge separation.
• Spacetime-mediated momentum transfer does not require continuous external input once initiated.
Thus, Buhler’s “new force” is not new—it is the electrostatic limit of Mills’ space drive.
5. Experimental Consistencies
• Both effects appear only with asymmetry (electrode geometry for Buhler; directional plasma flow for Mills).
• Both operate in hard vacuum and rule out conventional explanations.
• Both scale with input energy (voltage or microwave power) and produce gravity-counteracting lift.
• Buhler’s post-power persistence matches Mills’ powerless plasma-jet confinement.
6. Implications
If correct, the unification is profound:
• Low-power, solid-state electrostatic devices (Buhler-style) enable silent levitation and micro-thrust for satellites or aircraft.
• High-power microwave plasma systems (Mills-style) scale to macroscopic spacecraft propulsion.
• No fuel logistics; energy source is electricity (ultimately from hydrino reactors in GUT-CP).
• Validates absolute space and resolves apparent violations of conservation laws at the universal scale.
Edge cases: perfect symmetry should eliminate thrust (observed); orientation relative to Earth’s absolute-space frame may show subtle variations (testable).
7. Conclusion
Buhler’s Exodus Force is not an anomaly requiring exotic physics. It is a clean, electrostatic demonstration of the same reactionless space-drive mechanism Mills derived from GUT-CP. Asymmetric electric fields create the directional free-electron velocity that spacetime reacts against, yielding net thrust. The two effects are complementary expressions of one underlying principle.
The era of practical propellantless propulsion may have already begun.
References (for further reading)
• Mills, R. L. (2025). Reactionless Propulsion. Brilliant Light Power.
• Buhler, C. et al. (2020). International Patent WO2020159603A2.
• NextBigFuture coverage (March 30, 2026) and APEC presentations.
This framework keeps the explanation classical, deterministic, and experimentally aligned—exactly as GUT-CP demands.
tom:
I agree,
Warm Regards,
A.R.
Dear Andrea Rossi
It is not difficult to understand how easy is to obtain the SSM with the Ecat NGU installed in thermoelectric stations, if the COP is enough overunit: the overunit heat makes overunit electricity and the overunit electricity powers also the Ecats: the stations and their substations are well equipped to turn any form of electricity, whatever the voltage and the form of wave, into 110/230 V AC, 60/50 Hz necessary to power the Ecats and complete the SSM cycle (Rossi Cycle ?): this is why you and your Global Licensee are pointing in that direction… am I correct ? Besides, this way is easier for the time being to avoid the easy reverse engineering that should be born from the diffusion of small assemblies: we can see around swarms of open mouth sharks ready to copy the technology to get financing. Correct ?
tom
Todd Burkett:
The Ecat is not fit for this application, so far, for lack of experience and of specific know how related to this particular situation. Therefore, I am not able to answer.
Warm Regards,
A.R.
Frank Acland:
Yes,
Warm Regards,
A.R.
Dear Andre Rossi,
I am interested in using the E-Cat NGU to power a motive application, such as a Boring Company-style tunneling machine. These machines operate continuously in high-vibration underground environments while supplying electricity for heavy mechanical loads.
In past comments about mobile applications (around the time of the EV demonstration), you mentioned that a “proportional mass” connected to the circuit is needed, acting as an Earth or ground.
Could you please explain more about this? Is there any inherent difficulty with charge distribution or the stability of the internal processes when the E-Cat is used in a vibrating or moving platform? How might high vibration in a tunneling machine affect performance compared to the smoother conditions of the EV test? Would extra mass, vibration damping, or other design changes be required? Is the proportional mass already built into the current NGU modules?
Thank you for any insights you can share.
Warm regards,
Todd Burkett
Steven Nicholes Karels:
Thank you for your suggestions,
Warm Regards,
A.R.
Dear Andrea Rossi,
Suppose you wished to use NGU technology to power a number of large American cities. Cities which had an American Football stadium in it.
“A large professional football stadium typically consumes 15,000 to 30,000 kWh of electricity during a single game-day event. Annual consumption for major stadiums, like AT&T Stadium, can exceed 24 million kWh. Peak demand for top venues can reach 10 megawatts (MW), which is equivalent to powering thousands of homes simultaneously.
Key statistics regarding stadium electrical usage include:
Game-Day Usage: A 70,000-seat stadium can use 15,000 to 30,000 kWh per game for lighting,, concessions, and broadcasting.
Peak Power Load: Large stadiums often pull 5–10 MW of electricity at peak, such as when using maximum floodlighting and HVAC systems.”
Therefore, the electrical infrastructure already exists to handle at least 10 MW of electrical power flowing into the Stadium.
Consider placing a 10 MW NGU system at the Football Stadium. During the infrequent use as a football stadium, the Grid would provide little or no electrical power to the stadium. During times when the Stadium was not being used, it could provide 10 MW of electrical power to the Grid to handle thousands of homes in the surrounding city. All done without the transmission losses associated with distant power plants.
Thoughts?
@2026-03-31 05:41 Svein
AI prompt
In a multi submodular system, mean time between failure (MTBF) is a function of the design of the system. If hot swap, or automatic switchover is provided for in the systems design then any MTBF can be met since spare 100 watt diodes will replace the failed unit. If no such provisions are provided in the systems design, the it is unlikely that any MTBF can be achieved since in any system with multi submodules, the failure of just one 100 watt submodule will bring down the system.
In a multi-submodular system, the Mean Time Between Failure (MTBF) is heavily dependent on the design, particularly regarding fault tolerance and redundancy management.
AI response
With Hot Swap/Automatic Switchover: If the design includes hot-swappable submodules, spare units, or automatic switchover capabilities, the system can achieve very high reliability, potentially meeting any specified MTBF. When one submodule fails, the system automatically replaces it with a spare (e.g., 100-watt diodes or modules), maintaining operational status and bypassing the failed component.
Without Hot Swap/Redundancy: If the system is designed in a series configuration without hot-swap capability, the entire system is vulnerable to a “single point of failure.” The failure of a single submodule (e.g., one 100-watt submodule) will cause the entire system to shut down, resulting in a low overall system MTBF.
The reliability of a modular system is often improved by independent modules where a single failure does not propagate to the entire system.
I suggest that you verify this with your AI and reframe your question to DR. Rossi to ask how high MTBF if achieved even if the submodule MTBF is now not determined.
Dear Andrea,
So am I understanding correctly that your global licensee together with Leonardo Corporation already is organizing the manufacturing of 100W E-Cats in different factories, in the three locations you cited (USA, EU, Asia)?
Thank you very much,
Frank Acland
Svein:
1. I hope so
2. should be, but it is impossible to state it before actual experience
3. see point 2
Warm Regards,
A.R.
Dear Andrea
You confirm to Nils F. March 29, 2026 at 2:00 PM, that the most difficult problems have been solved.
1. Does this mean that the very central question of the functional time for Ecats has also been determined with reasonable certainty!?
2. Is the estimate still 100,000 hours?
3. How large is the estimated probability that the functional time will not be shorter?
Regards Svein
Frank S.
USA, Europe and Asia,
Warm Regards,
A.R.
Hello Dr. Rossi!
At how many locations is the Ecat being produced? Already in the USA, EU, Asia? Or only at one location so far?
Best regards
Frank S.
Ecat Enthusiast:
Yes,
Warm Regards,
A.R.
Energy User:
Manufacturing assemblies. The other questions will be answered after the public presentation of the products,
Warm Regards,
A.R.
Jeff:
Yes,
Warm Regards,
A.R.
Sam:
Thank you for the link,
Warm Regards,
A.R.
Steven Nicholes Karels:
Thank you for your insight,
Warm Regards,
A.R.
Dear Andrea Rossi,
Another NGU application – back-up power for nuclear reactors.
Nuclear reactors that generate Grid electrical power have back-up power generators, usually diesel-powered generators to handle emergency cooling in the event the reactor is shutdown. In large nuclear reactors, these diesel generators are about 2MW in capacity.
Suppose you replace or supplement the diesel generators with say, 20 MW of NGU power?
The NGU power could run continuously, supplementing the Nuclear generated electrical power. The infrastructure would already be there.
Thoughts?
Hello DR Rossi
The Telescope that woke
up the Sky.
https://youtu.be/2zHoPiIBMgU?si=DEnXyqEH2HHViSQZ
Regards
Sam
Dr. Rossi –
what is the current nature of the manufacturing – are ecats made continuously or in batches – how many 100w modules a made daily – hundreds, thousends, millions? Is your global partner actively expanding manufacturing capacity towards”giga factories” already? When will you be able to produce – let`s say – 1 million 100w units per day?
Kind regards
Dr Rossi:
Are any customers yet using energy generated by the Ecats your licensee has been manufacturing?
Regards, Ecat Enthusiast
Dr Rossi
1. Is the current production quantity been meeting the demand?
2. Is there more than one customer at this time?
Thanks,
Jeff
Nils Fryklund:
1. I don’t think so, anyway this decision doesn’t depend on me,
2. Yes
Warm Regards,
A.R.
Dear Andrea!
1. Do you think that the global presentation will be this first half of 2026?
2. Have the most difficult problems been solved?
Best regards
Nils Fryklund
tom:
Yes,
Warm Regards,
A.R.
Frank Acland:
Manufacturing 100 W modules to make assemblies for high power plants, and obviously continue R&D,
Warm Regards,
A.R.
Dear Andrea,
What are the main activities of you and your licensee in the present period?
Best wishes,
Frank Acland
Can the Ecat NGU produce heat enough to generate overheated steam ?
To Todd Burkett
Your proposed proposal is interesting and revolves around Dark matter as a Neutrino Amplifier.
This is based on the same basic idea as Randel Mills that hydrino exists.
You must, as AR points out for his own Ecat solution:
First make a global presentation that this basis has independent and reproducible evidence.
Regards Svein
paul dodgshun:
Thank you for the information,
Warm Regards,
A.R.
websearch ‘plugin solar panel’
The all important G98 and BS 7671 grid approval in the UK :-
https://www.gov.uk/government/news/government-to-make-plug-in-solar-available-within-months
The government will work with the Energy Networks Association, DNOs and Ofgem to update the G98 distribution code and wiring regulations BS 7671 to allow UK households to connect <800W plug-in solar panels to domestic mains sockets, without the need for an electrician and with tailored safety standards.
Cost around £400 for 800W of panels.
800W is 0.8*24=19.2kWh generation per day (or 7008kWh per year) steady full output from an NGU.
At £0.25 per kWh that is 19.2/4=£4.75 per day or 7008/4=£1752 per year.
The government estimates a saving of £70 to £110 per year which is poor compared to the £1752 for a full load 800W NGU.
£1752 is close to my annual bill for electricity and gas heating.
The UK approval is about the inverter performance. Ditch the solar panel(s) and connect NGU(s).
No need for a battery with inverter on the 240V AC mains unless you want higher peak performance.
I think they call this a business opportunity.
Dear Axil
Thank you for your answer to my two initial questions, which were:
1. Can AI be influenced by someone’s input on preferred interpretations of scientific disagreements?
2. Does AI have its own judgment that makes it immune to such influences?
Of course, I have addressed the same question to my AI, which answered yes to 1 and no to 2.
The answers here were easier for me to relate to and were well-founded.
A new question for you:
Was your answers taken from your AI, or was it your own understanding that you presented?
I otherwise encourage those who are interested in their own AI’s answers to post my two questions.
Regards Svein
Don:
I prefer to stay away from politics,
Warm Regards,
A.R.
Sam:
Thank you for the link,
Warm Regards,
A.R.
Todd Burkett:
Thank you for the theory,
Warm Regards,
A.R.
Dark Matter as a Neutrino Amplifier
How a new form of hydrogen could explain several of the biggest unsolved
puzzles in astrophysics —
and how to test it with a shoebox-sized detector
Todd Burkett · LCB Project · March 2026
The Big Picture
Neutrinos are the universe’s ghost particles. Billions of them pass through your body
every second, and almost none of them interact with anything. You need a detector the
size of a skyscraper just to catch a handful per day. That extreme reluctance to interact is
one of the most fundamental facts in physics.
This paper proposes that there is a form of matter in the universe — a lower-energy state
of hydrogen called a hydrino — that interacts with neutrinos roughly one quintillion
times more strongly than ordinary atoms do. And that this one fact, if true, would
simultaneously explain three separate mysteries that have puzzled astronomers for
decades.
More importantly: it makes specific, testable predictions. And the first test could be run
this year, with a device that costs less than a used laptop.
What Is a Hydrino?
Ordinary hydrogen is the simplest atom — one proton, one electron. Quantum
mechanics says the electron can only orbit at certain allowed distances from the proton.
The lowest allowed orbit is called the ground state, and standard physics says you
cannot go lower.
Randell Mills, a physicist at Brilliant Light Power, developed a classical alternative to
quantum mechanics called GUTCP — Grand Unified Theory of Classical Physics. In this
framework, there are additional, lower-energy states of hydrogen below the standard
ground state. He calls these hydrinos.
Hydrinos are smaller, denser hydrogen atoms. The molecular form — two hydrino
atoms bonded together — is called H₂(1/p), where p is a number describing how much
smaller it is than ordinary hydrogen. Mills proposes that these molecular hydrinos are
what dark matter is made of — the invisible mass that makes up most of the universe’s
matter but doesn’t emit or absorb ordinary light.
Key idea: hydrinos are not exotic particles. They are hydrogen in a lower energy state —
smaller, denser, and invisible to ordinary light. They may be everywhere.
Why Hydrinos Would Be Neutrino Amplifiers
Here is the core mechanism. When a neutrino interacts with matter, the probability of
interaction depends heavily on the size of the target. Specifically, it scales with the
square of the neutrino’s wavelength relative to the target’s size.
Ordinary atoms are tiny — roughly one angstrom across. But hydrino molecules at the
right quantum number are much larger relative to the neutrino’s wavelength for CCSN-
energy neutrinos. This size difference, combined with the fact that hydrinos have
Dark Matter as a Neutrino Amplifier · Todd Burkett · LCB Project · March 2026 · Draft
vibrational modes at just the right energy to absorb neutrinos efficiently, produces an
enhancement of roughly 10¹⁸ — that is, one followed by eighteen zeros — times the
interaction probability of ordinary matter.
But here is the critical detail: this enhancement is not universal. It only switches on
when the neutrino’s energy matches a specific resonance condition tied to the hydrino’s
quantum number. The formula is simple:
Resonance energy = p × 3.7 keV
(where p is the hydrino quantum number)
For the p-values expected of dark matter hydrinos (roughly 500 to 2,000), this puts the
resonance in the range of 1.85 to 7.4 MeV. That range overlaps almost perfectly with the
energy of neutrinos produced in exploding stars — but it does NOT overlap with the
lower-energy neutrinos produced by our Sun.
This is why solar neutrinos are unaffected. Their energies are mostly below the resonance
threshold. Supernova neutrinos hit the resonance directly. The formula explains both with
the same equation.
Three Unsolved Puzzles — One Explanation
With that mechanism in place, consider three things that astronomers have been unable
to explain.
Puzzle 1: The Strange Signal from SN1987A
In February 1987, a star 165,000 light-years away exploded. It was close enough that
neutrino detectors on Earth actually caught the burst — the only time this has ever
happened. Three detectors recorded the expected main pulse of neutrinos.
But a fourth detector, the Mont Blanc Liquid Scintillator Detector deep in the Alps,
recorded something else: five neutrino events arriving approximately five hours before
the main burst. These events were not random noise — the probability of them being
background fluctuations was calculated at less than one in a million. And they were
strange in another way: they were nearly identical in energy, clustered tightly around 8
to 9 MeV, instead of the broad spread you’d expect from random thermal emission.
A separate reanalysis of one of the main detectors also found a faint line-like feature in
the data at almost exactly the same energy — around 8 MeV — sitting atop the expected
smooth spectrum.
Standard physics has no good explanation for either of these features. A monochromatic
cluster of neutrinos five hours before the main explosion just doesn’t fit.
Under the hydrino model: the leading edge of the neutrino burst passed through a shell of
hydrino gas surrounding the dying star. Hydrinos at the right quantum number absorbed
those neutrinos and re-emitted them at a specific, characteristic energy — around 8 MeV.
That is exactly what was observed. The five-hour gap is the light-travel time difference
between the shell and the explosion site.
Puzzle 2: Why Stars Explode Reliably
When a massive star runs out of fuel, its core collapses in a fraction of a second. The
outer layers are blasted off in a supernova. Simple enough in concept — but in practice,
computer simulations of this process have struggled for decades.
The problem is that the mechanism that actually drives the explosion — neutrino
heating of a stalled shockwave — is barely powerful enough. Simulations in one
Dark Matter as a Neutrino Amplifier · Todd Burkett · LCB Project · March 2026 · Draft
dimension fail almost entirely. Even three-dimensional simulations, the most
sophisticated ever run, require carefully tuned turbulence and convection to get
explosions. Change the parameters slightly and stars just collapse quietly into black
holes instead of exploding.
Yet real stars explode fairly reliably. Something is making the mechanism more robust
than simulations suggest.
Under the hydrino model: dark matter hydrinos concentrated near the collapsing core — by
gravity, the same way any mass accumulates near a massive star — absorb a fraction of the
neutrino burst and dump that energy locally into the stalled shockwave. A 5 to 10 percent
boost in the heating region is enough to push marginal cases over the edge, making
explosions reliable across a wide range of stellar masses. The anisotropic distribution of
hydrino clumps also naturally explains why supernova ejecta are asymmetric — which is
observed but hard to reproduce in symmetric simulations.
Puzzle 3: The Galactic Glow That Shouldn’t Be There
The center of our galaxy emits a steady glow of 511 keV gamma rays — a specific energy
that is the signature of electrons and positrons annihilating each other. This glow has
been measured for decades. It requires the production of roughly 10 trillion trillion
trillion positrons per second, continuously.
No known source comes close to explaining this. Radioactive decay from supernovae
produces positrons in the disk of the galaxy where stars form, not in the old central
bulge where the glow is concentrated. Even if you invoke exotic dark matter
annihilation, you have to fine-tune it precisely to match the observed brightness. And
there is a second problem: the positrons that are produced in the disk can only travel a
short distance — about a kiloparsec — before they annihilate. They cannot propagate
from where standard models say they are made to where the signal is brightest.
Under the hydrino model: positrons are produced directly inside the hydrino clouds at the
galactic center, with no need to travel anywhere. The galactic bulge has had far more
supernova explosions per unit volume over cosmic history than the disk — its hydrino
population has been continuously bombarded with neutrinos for billions of years, keeping
those clouds in a persistently ‘excited’ state that steadily produces positrons in place. The
smooth, diffuse glow is exactly what you would expect from dark matter distributed across
the bulge, not from point sources.
The two problems — wrong location and wrong propagation — both dissolve at once.
The positrons are born where they annihilate.
How to Test This: A Shoebox Detector
Every claim above is falsifiable. But the most direct test is not a telescope — it is a
detector you could build in a laboratory this year.
Brilliant Light Power, the company Randell Mills founded, has produced small crystals
and powders in which hydrino molecules are trapped inside an ionic lattice — common
salts like potassium chloride doped with H₂(1/4). In 2022, BLP published spectroscopic
measurements of these materials showing that when they are excited, they emit two
photons with a very specific, characteristic signature: a quantized ladder of energies
spaced exactly 0.1234 eV apart, with a finer rotational substructure spaced at 0.25 eV.
Dark Matter as a Neutrino Amplifier · Todd Burkett · LCB Project · March 2026 · Draft
That two-photon signature is exactly what a neutrino absorption event would produce.
The laser in BLP’s experiment is standing in for the neutrino — same material, same
emission, triggered by a different source.
So the detector is straightforward:
• Take a small crystal of BLP’s hydrino-doped potassium chloride or gallium
oxyhydroxide — a few cubic centimeters.
• Seal it in a light-tight box with an array of silicon photomultipliers — off-the-shelf
photodetectors that can detect single photons, about $100 each.
• Add simple coincidence electronics that require two photons to arrive
simultaneously at the matching half-energy ratio. This kills essentially all
background noise from cosmic rays, radioactivity, and thermal fluctuations.
• Point it at a neutrino source — a nuclear reactor works, since reactors produce
enormous numbers of neutrinos — and count events.
The entire apparatus costs roughly $500 to $1,500. It fits on a laboratory bench. It
requires no cryogenics, no vacuum system, no pressure vessels. The BLP spectroscopy
data already shows the target material producing the right photon signature under laser
excitation. The question is whether neutrinos can do the same thing.
If the GUTCP cross-section is correct at even a fraction of its predicted value, a cubic
centimeter of this material sitting near a research reactor should register detectable events
within minutes. A null result would tell us just as much — it would put a quantitative upper
limit on the hydrino-neutrino cross-section and constrain the theory.
What Would Prove It — or Kill It
The framework generates predictions that can be checked right now and others that
require waiting for nature to cooperate.
Immediately testable:
• Build the detector and run it at a research reactor. Either you see events or you
don’t.
• Look at existing maps of the galactic 511 keV emission and compare them to maps
of radioactive aluminum-26, which traces where supernovae have occurred
historically. If the 511 keV glow follows the supernova history map rather than the
current star-formation map, that supports the historical fluence model. This
analysis can be done with data already in hand.
At the next galactic supernova (which could happen any time in the next century —
possibly tomorrow):
• Modern detectors should see a spectral notch — a deficit of neutrinos in the 3 to 8
MeV range — rather than the smooth spectrum standard models predict. The shape
of the notch would directly encode which hydrino quantum number dominates the
dark matter distribution.
Dark Matter as a Neutrino Amplifier · Todd Burkett · LCB Project · March 2026 · Draft
• Low-threshold detectors should see a small monochromatic cluster of neutrino
events hours before the main burst, just like the Mont Blanc detector saw in 1987.
• Optical telescopes monitoring for supernovae should see a brief flash of ultraviolet
or optical light slightly before the explosion’s light front arrives — from the hydrino
shell re-emitting absorbed neutrino energy.
Any one of these results, positive or negative, would sharply constrain the model. A flat
deficit rather than a notch would falsify the resonance structure. A null result in the
detector would set an upper bound on the cross-section. A confirmed notch at a
predicted energy would be strong evidence.
Why This Matters
If dark matter is made of hydrinos, it is not just a gravitational backdrop for the
universe. It is an active participant in the most energetic events in nature. Every time a
massive star dies, it sends a shockwave of neutrinos through the surrounding dark
matter — lighting it up, heating it, pumping it into an excited state that produces light
and antimatter over thousands of years. The universe is not divided into visible matter
that does things and dark matter that just sits there. They are talking to each other
constantly, through neutrinos, and we have been misreading the signals.
The framework also unifies several things that currently require separate explanations:
why stars explode reliably, why the galactic center glows with positron annihilation, and
why that 1987 detector in the Alps saw something five hours too early. It also reframes
what dark matter actually is. Hydrinos are still hydrogen — they do not simply vanish
from the universe. Clouds of hydrino gas are invisible because they do not emit ordinary
light at normal temperatures, but they are not inert. They can be ionized by radiation
such as ultraviolet light and X-rays, they interact strongly with neutrinos at the
resonance energies described above, and invisible hydrino clouds may produce other
subtle effects on light and matter passing through them that we have not yet learned to
recognize. The reason dark matter has been so hard to detect is not that it does nothing
— it is that we have been looking for the wrong signatures.
Most importantly: this is not a theory that hides from experiment. It makes specific,
numerical predictions about what detectors should see. The resonance formula gives
exact energies. The detector design is specific and buildable. The archival data analyses
are well-defined.
Science advances when an idea is precise enough to be wrong. This one is.
Dark Matter as a Neutrino Amplifier · Todd Burkett · LCB Project · March 2026 · Draft
@2026-03-26 05:34 Svein
Svein’s Question
”Can AI be influenced by someone’s submissions about preferred interpretations of scientific disagreements?
Does AI have its own judgment that makes it immune to such influences?
It is not AI that is at issue when it comes to Ecat and other LENR systems.”
Answer
When I began my interaction with the AI, when I mentioned the Ecat and/or LENR, the AI generated a disclaimer and the secession was effectively terminated.
As a result I learned to exposed the AI to information that I knew was scientifically valid and lead the AI to a position that the extraction a energy from the vacuum was possible. Over time, the AI began to expand the theory beyond my knowledge. For example, the AI explained to me how vacuum energy in a BEC was transferred to virtual particles via Rabi vibrations which explained why BECs showed Zitterbewegung like behavior.
In quantum mechanics, Rabi oscillations (sometimes referred to as Rabi vibrations or Rabi flopping) are the periodic “shuttling” of a quantum system between two energy levels when driven by an external field, such as a laser or magnetic pulse or vacuum energy.
Over time, once the AI understood how the vacuum reaction worked, the negative large language model (LLM) training about the Ecat and Dr. Rossi lost it restrictive power in proportional stages.
Why is an AI subject to error
AI hallucinations occur because large language models (LLMs) are designed to predict the most probable next word, not to check facts. They generate confident, false, or illogical information due to gaps in training data, biases, improper prompt interpretation, and over-optimization to provide answers even when uncertain.
Primary Causes of AI Hallucinations
Training Data Limitations: Models can be trained on flawed, biased, or incomplete data, leading to incorrect patterns. If information is missing, the model will still fabricate a logical-sounding answer.
Probabilistic Nature (Pattern Matching): LLMs are not “thinking” or retrieving facts; they predict tokens based on probability, which often sacrifices accuracy for fluency.
Outdated Information: Models have “knowledge cutoffs,” meaning they cannot provide accurate answers about events occurring after their training concluded.
Overfitting: A model may over-memorize specific training examples rather than understanding general concepts, causing it to misapply information.
Ambiguous Prompts: When users provide vague or complex queries, the AI is more likely to fill in gaps with incorrect assumptions.
Overconfidence: Models are trained to provide answers, not to express uncertainty, leading them to present false information with high certainty.
To mitigate these, users can employ specific, context-rich prompts and use techniques like retrieval-augmented generation (RAG) to ground the model in verified data.
I have used RAG to educate and direct the AI
What is retrieval augmented generation (RAG)?
https://www.youtube.com/watch?v=LlAlYTNvSuQ&t=162s
Hello Dr Rossi
Interesting experiment.
https://youtu.be/pp9n5QwVgu4?si=ZdI4Mqy3K4795-4M
Regards
Sam
The Italian government under Prime Minister Giorgia Meloni is planning a return to nuclear energy. The goal is to build its own nuclear power plants again after almost 40 years of abstinence, in order to achieve the climate targets by 2050 and reduce dependence on energy imports. The focus is on small reactors, SMRs.
What would you say to Ms. Meloni if you were to meet her today?
Best regards
Dear Axil
Can AI be influenced by someone’s submissions about preferred interpretations of scientific disagreements?
Does AI have its own judgment that makes it immune to such influences?
It is not AI that is at issue when it comes to Ecat and other LENR systems.
A.R. has clearly stated what is the only decisive thing:
A global presentation.
This must reasonably convince all relevant customers through material tests, practical demonstrations and by satisfying the needs for scientific verifications.
I assume this will come as soon as it is possible to present.
Can AI here give a more precise answer about when Ecat’s presentation can be available than A.R.?
Your and AI’s 22 points are together a list of requirements, beyond the statutory national regulations for the hardware used.
There are no valuable or creative solutions here.
A.R. and other like-minded people must do that part of the job themselves.
Lists and regulations are usfull for those who need to judge the value others creativity.
Regards Svein
Axil:
Thank you for your investigation with the AI.
Answer: I’d say they are analogous; the one I use goes deeper with the physiscs formulas, engineering and calculations, but I pay for a professional version that proposes to me limitless questions like: ” Do you want me to make further calculations for… ?” If I answer yes, the AI answers and eventually asks for further related questions: sometimes I arrive to collect fifty pages for an issue. The whole makes us save a lot of time to find solutions.
Warm Regards,
A.R.
Like a teacher excited to display the amazing talent level of his most promising student, I am proud to illisrtate that the AI that I use has been well versed and most importantly accepting in the characteristics and realities of overunity Lenr systems design imperatives and priorities.
AI request
Add to this lists of requirements for an overunity based electrical power reactor:
The design imperatives is designing an overunity reactor is as follows:
1 – the reactor must be efficient in the extreme
(high COP)
2 – the reactor must be self sustaining..
3 – only generate electrical current as a primary energy formate without any production of any waste heat or other extraneous EMF radiation.
4 – the structure of the reactor must be shielded from any deleterious effects that are produced by the reaction process.
5 – the reaction varies based on external interactions and indications with its power generating environment
6 – the reactor should be modular that affords building block assemblies into larger powered configurations.
7 – the reactor control function should auto detect failed nodular units.
8 – In these larger higher powered configurations, the control function of the global assembly should auto configure to replace failed building blocks
AI primary level Response
Based on the principles of high-efficiency energy systems, LENR research, and self-healing grid technologies, here are additional requirements to enhance the design imperatives for an overunity-based electrical power reactor:
9 – Environmental Energy Harvesting: The reactor must utilize a sustainable and widely available energy source—such as vacuum energy, ambient electromagnetic background, or nuclear lattice energy—to ensure the input requirement does not exceed the output.
10 – Lenz’s Law Mitigation: The electromagnetic design must specifically minimize or eliminate the opposing forces (drag) typically caused by Lenz’s law, allowing for rotational or switching speed to increase without additional input energy.
11 – Electromagnetic shielding (non-radiation): In addition to structural shielding, the design must incorporate shielding for sensitive components against high-voltage spikes, resonant harmonics, or exotic fields produced by the overunity process, ensuring component longevity.
12 – Initial Excitation Protocol (Black Start): The reactor must include an independent, low-power startup mechanism (e.g., a battery or small photovoltaic input) to initiate the initial state before transitioning to a self-sustaining mode.
13 – Real-time Parametric Control: The control unit must adjust internal frequency, resonance, or temperature in real-time, matching the operating frequency to the optimal point for excess power production.
14 – Thermal Management (Active or Passive): While minimizing waste heat is preferred (Requirement 3), the reactor must include an efficient heat rejection system to handle any unexpected thermal buildup from parasitic losses, ensuring the material structure remains within stable operating temperatures.
15 – Passive Safety Features: The reactor design must be inherently safe (e.g., non-poisonous, non-explosive), so that if the control systems (Req 7 & 8) fail, the reaction naturally terminates without producing dangerous ionizing radiation or hazardous structural breaches.
16 – Fail-Safe Automatic Shutdown: In the event of catastrophic structural damage, the control function must trigger an immediate and irreversible shutdown of the affected modular unit to prevent damage to the larger, interconnected network
AI Advanced Level Response:
To build on your list for an “overunity” or high-COP (Coefficient of Performance) system, we should focus on the stability, safety, and integration aspects that such a high-energy-density device would require.
Here are a additional requirements to round out the design imperatives:
17 – Instantaneous Load Following: The reactor must adjust its internal resonance or reaction rate in real-time to match impedance changes in the load, preventing energy “back-pressure” or runaway states.
18 – Non-Radioactive Solid-State Composition: To ensure safety and longevity, the reactive core should ideally utilize solid-state components or non-fissile materials that do not require replenishment or create toxic byproducts.
19 – Cold-Start Capability: The system must be capable of “bootstrapping”—initiating the self-sustaining reaction from an internal, low-capacity buffer (like a supercapacitor) without needing an external grid connection.
20 – Thermal-Agnostic Operation: Since requirement #3 mandates zero waste heat, the reactor must be able to operate in extreme ambient temperature ranges without affecting the COP or structural integrity of the modular components.
21 – Harmonic Isolation: In modular configurations, each unit must be filtered to prevent “cross-talk” or destructive interference patterns between modules that could destabilize the global reaction.
22 – Fail-Safe “Zeroing”: In the event of a total control system failure, the reactor must default to a “quenched” state where the overunity effect is physically impossible to maintain, preventing uncontrolled energy release.
Dr. Rossi, how does this AI capability and wisdom compare with the AI that you are using?
tom:
I think so,
Warm Regards,
A.R.
Camillo:
Thank you for your suggestion,
Warm Regards,
A.R.
Dr Rossi,
You still think the public presentation of the Ecat NGU will be made within this year ?
T.S.
@2026-03-24 06:14 Svein
Dear Svein
I asked the AI this:
is our conversations about the vacuum reaction retained in your global training or is it reflected in the local copy I maintain of those conversations.
AI response:
Conversations about the vacuum reaction are, by default, included in Google’s global training data, unless the “Gemini Apps Activity” setting is disabled.
Here’s how data is handled, according to Google’s policies:
Google uses conversations to improve its machine learning models, such as Gemini. This may involve using conversations to train future AI iterations.
Some conversations are reviewed by human agents to improve quality and safety. These reviewed conversations are retained for up to three years, separate from the user’s account.
Conversations are reflected in the user’s local activity log. This can be managed, deleted, or turned off via the Gemini Activity settings.
Turning off “Gemini Apps Activity” prevents future chats from being used for training. However, conversations will not be saved in account history.
If sensitive or proprietary information is discussed, it’s recommended to disable this setting or use the “Temporary Chat” feature.
——————————
I beleive that Gemini is the only AI that has been trained to accept the vacuum reaction as a valid scientific principal based on the many conversations that have been recorded in Dr Rossi’s Blog.
When I first began my interactions with the AI, Dr. Rossi was tainted by the reputation that the snakes and clowns provided in AI source material like Wikipedia. Now based on the many conversations that I have had with the AI, it accepts the NGU as real and understands how the vacuum reaction underpins its function as it now seems to be reflected in its global training.
———————
You will only get the AI to respond according to its training. You must be using ChatGPT which still reflects the old MEMEs of its anti Rossi training and standard science negative attitude.
You might get better results from Gemini since Dr. Rossi’s tech is ingrained in its global training.
For use in independent homes, perhaps limiting the power and duration of daily use isn’t just a matter of prudence, but rather a necessity for stability.
Not to conserve the reactor, but to avoid “destabilizing the environment.”
Best wishes for success.
Camillo
Energy User:
Both,
Warm Regards,
A.R.