I did not start RF Safe because I thought the world needed another technology website. I started it because of a promise.
In 1995, my daughter Angel Leigh Coates died from a neural tube defect. That loss changed the entire trajectory of my life. At the time, the clearest actionable science around neural tube defects pointed to folate, B12, and early pregnancy nutrition. So before RF Safe existed, before the FDA’s 1998 folic acid fortification mandate was fully implemented, I was already spending my own money trying to warn and educate women about folic acid, B vitamins, and neural tube defect prevention.
But Angel’s case did not fit neatly into the explanation I was given. Her mother’s folate status was not the obvious failure point. The timing, the occupational electromagnetic exposure, and then the 1997 chick embryo study by Farrell and colleagues forced me to ask a harder question.
What if the real problem was not simply a missing nutrient?
What if biology had the parts — but the instructions were degraded?
That is the heart of the low-fidelity biology hypothesis.
The Wrong Question: “What Disease Do EMFs Cause?”
For decades, the debate around electromagnetic fields has been trapped inside a flawed question:
Do EMFs cause cancer?
Do EMFs cause autism?
Do EMFs cause neural tube defects?
Do EMFs cause disease X?
That framing assumes biology works like a courtroom: one exposure, one crime, one verdict.
But embryogenesis, neurodevelopment, immunity, metabolism, aging, and tissue repair do not work that way. They are not single-output machines. They are high-dimensional, adaptive, timing-sensitive biological control systems.
So the better question is not:
“Which disease do EMFs cause?”
The better question is:
“Can invisible environmental fields lower the fidelity with which biology executes its own programs?”
That is a fundamentally different scientific question.
If the answer is yes, then we should not expect EMFs to map cleanly to one disease. We should expect them to behave like upstream noise in a biological operating system. The final disease label would depend on timing, genetics, nutrition, redox state, infection, drugs, heat, toxicants, repair capacity, and developmental stage.
That is not a weakness of the hypothesis. That is the hypothesis.
Low-Fidelity Biology: A Definition
By biological fidelity, I mean the precision with which cells interpret signals, maintain timing, and execute complex programs.
During early development, cells are not just dividing randomly. They are coordinating membrane voltages, calcium waves, ion-channel gating, mitochondrial redox states, chromatin marks, methylation patterns, cytoskeletal forces, migration, proliferation, and differentiation.
Development is not merely chemistry. It is timed, structured, bioelectric computation.
Modern developmental biology already recognizes that bioelectrical signals controlled by ion channels contribute to communication during development, and that membrane voltage and ion-channel activity help regulate cell proliferation, differentiation, and patterning. Michael Levin’s work has gone further, describing bioelectrical networks as systems that process morphogenetic information and influence gene expression and large-scale anatomical outcomes.
The ceLLM framework pushes this into a broader language: cells are not passive chemical bags; they are adaptive, inference-making, electrically sensitive systems. In the material you shared, folate and B12 are framed as part of the “hardware supply chain” for methylation and developmental tuning, while pulsed EMFs are framed as entropic static that can degrade cellular computational fidelity.
That is the core idea:
A healthy cell is not just alive. It is computing with high fidelity.
A low-fidelity cell may still survive. It may still divide. It may still compensate. But it does so with more noise, more error correction, more epigenetic patching, more oxidative stress, and more mistimed signaling.
Eventually, that low-fidelity state can become a meta-disease state.
Not one disease.
A biological condition that makes many diseases easier.
The Meta-Disease State
A meta-disease state is not a diagnosis. It is a loss of biological signal quality.
It means the system’s resilience has dropped.
It means the cell is spending more energy correcting noise than executing its native program. It means calcium timing is less clean. Membrane potentials are less stable. Mitochondrial signaling is less coherent. Chromatin marks are more reactive. Methylation becomes less like precise tuning and more like emergency patchwork.
This is why the search for a simple EMF-to-disease map may keep failing.
A weak, chronic, timing-based environmental stressor may not behave like a poison. It may behave more like jitter in a clock, static in a communication channel, or corrupted timing in a control system.
When the system is simple, noise may be tolerated.
When the system is running the most complex program in biology — building a human embryo — noise can matter enormously.
Neural Tube Closure: The Ultimate Fidelity Test
Neural tube closure is one of the most unforgiving events in human development.
The CDC notes that neural tube defects develop very early in pregnancy, often before a woman knows she is pregnant, and that 400 mcg of folic acid before and during early pregnancy can help prevent neural tube defects. The USPSTF recommends that people planning to become pregnant or capable of pregnancy take 0.4 to 0.8 mg of folic acid daily, starting at least one month before conception and continuing through the first two to three months of pregnancy.
That public-health point matters. Folate biology matters. B12 matters. One-carbon metabolism matters. DNA synthesis and methylation matter.
But folate does not close the neural tube by itself.
Folate supplies essential biochemical capacity. It helps stock the cellular construction site. But the embryo still needs accurate timing signals to tell the cells where to move, when to divide, when to change shape, when to fuse, and when to stop.
That is where the low-fidelity hypothesis becomes important.
If an embryo has the building materials but the bioelectric timing code is degraded, the outcome can still fail.
This may help explain why neural tube defects are not reducible to folate alone. The CDC itself lists several risk factors beyond low folate, including poorly controlled diabetes, certain medications, overheating, fever, and genetic or multifactorial influences.
The old model says: “Find the missing ingredient.”
The low-fidelity model says: “Also examine the integrity of the signaling environment.”
The 1997 Chick Embryo Signal
This is why Farrell et al. mattered so much to me.
In 1997, Farrell and colleagues reported that weak electromagnetic fields could induce morphological abnormalities in developing chick embryos. The PubMed summary describes the result directly: weak EMFs induced morphological abnormalities in developing embryos. Earlier ELF chick embryo work also reported sensitivity to extremely low-frequency, low-intensity fields and suggested that pulse shape could influence developmental effects.
To me, the meaning was not “this proves every human neural tube defect is caused by EMFs.”
That is not the point.
The point was this:
A developing embryo can be sensitive to weak, invisible electromagnetic structure during a high-fidelity morphogenetic window.
That is enough to demand better science.
Not dismissal. Not ridicule. Not “it cannot happen because it does not heat tissue.”
Better science.
The Thermal-Only Mistake
The thermal-only safety paradigm asks one narrow question:
Does this exposure heat tissue enough to cause damage?
But low-fidelity biology asks another question:
Can this exposure perturb timing, signal coherence, ion-channel behavior, calcium oscillation, mitochondrial redox signaling, or chromatin regulation without heating tissue?
Those are not the same question.
A radio signal does not need to cook tissue to interfere with timing. A field does not need to break DNA directly to change how cells read, fold, repair, or express DNA. In a timing-sensitive biological system, information disruption can matter even when bulk temperature does not change.
That is why newer mechanistic work is so important.
The 2026 Cell paper on an electromagnetic-field-inducible in vivo gene switch identified cytochrome b5 type B, CYB5B, as an essential mediator likely acting as an EMF sensor, and linked EMF-triggered gene-switch activation to rhythmic calcium oscillations rather than generic calcium influx.
That does not prove that everyday EMFs cause a named disease.
But it does undermine the lazy claim that weak electromagnetic fields are automatically biologically irrelevant.
It shows that biology can contain molecular hardware capable of coupling EMF input to calcium-timing biology under defined conditions.
That matters.
RF, Radial Glia, and Cortical Development
The same fidelity logic applies later in development.
If the hit occurs during neural tube closure, the concern is structural: the tube closes or it does not. But if the fidelity loss occurs later, during corticogenesis, synaptogenesis, glial maturation, myelination, or circuit pruning, the outcome may not be a gross birth defect. It may be altered wiring, altered differentiation, altered connectivity, altered sensory processing, altered behavior, or altered vulnerability.
That is where the RF cortical organoid work becomes relevant.
The NCBI GEO summary for the 2025 Cell Reports paper states that RF exposure in human cortical organoids regulated differentiation of human and non-human primate radial glia progenitors, maintained stem-cell identity, delayed differentiation, induced endogenous retrovirus expression in differentiated neurons, and that BET inhibitors rescued RF-induced developmental defects in human cortical organoids.
Again, this is not the same as saying “RF causes autism.”
The better interpretation is:
RF exposure has been reported to interact with early cortical developmental programs in organoid systems, especially pathways involving radial glia differentiation and chromatin-reading machinery.
That is exactly the kind of model-system signal one would expect under a low-fidelity biology framework.
Not one disease.
A perturbation of the developmental control layer.
Autism, Overgrowth, and Developmental Timing
The 2024 Courchesne, Muotri, and colleagues paper adds another piece to the timing puzzle. In brain cortical organoids derived from toddlers with autism, the researchers reported that organoid size was enlarged in ASD batches, that larger embryonic brain cortical organoid size correlated with more severe social symptoms, and that the largest ASD organoids showed accelerated neurogenesis. They also reported that NDEL1 activity correlated with growth rate and organoid size.
That study does not identify EMFs as the cause.
But it does support a critical principle:
Some autism-relevant biology is already present during embryonic development, in systems involving growth, neurogenesis, and developmental timing.
That matters because low-fidelity biology is a timing hypothesis.
The same upstream category — mistimed ion channels, perturbed membrane potentials, degraded calcium waveforms, mitochondrial noise, chromatin misreading, and epigenetic patching — could produce different downstream phenotypes depending on when the fidelity loss occurs.
Early enough, the concern may be neural tube closure.
Later, it may be cortical overgrowth, delayed differentiation, altered connectivity, or neurodevelopmental vulnerability.
Later still, it may appear as immune dysregulation, metabolic disease, infertility, neurodegeneration, cancer susceptibility, or accelerated aging.
The disease label changes.
The upstream fidelity problem may be shared.
Folate, Folic Acid, and the Bigger Point
There is a separate and important debate about folic acid, natural folate, methylfolate, folinic acid, fortification, consent, B12 masking, and population-wide exposure to synthetic nutrients. The transcript you shared emphasizes that distinction strongly, especially around synthetic folic acid, receptor competition, fortification policy, and informed consent.
But for this argument, the central point is not “folate good” or “folic acid bad.”
The central point is deeper:
Folate metabolism is part of biological fidelity.
The FDA required enriched cereal grain products in the United States to include 140 mcg folic acid per 100 g by 1998, and CDC has described folic acid fortification as associated with major reductions in neural tube defect prevalence. At the same time, CDC has acknowledged concerns that have been raised about high folic acid intake, including unmetabolized folic acid, B12-related issues, epigenetic effects, and cancer promotion, while stating that no conclusive evidence exists that recommended levels cause those conditions and that continued monitoring is needed.
That is exactly the kind of nuance the fidelity model requires.
Biology is not slogans.
Folate and B12 are not just “vitamins.” They participate in methylation, DNA synthesis, RNA synthesis, repair, and gene regulation. In ceLLM language, they help provide the physical supply chain for biological weight updates. But the supply chain is not the whole construction project.
You need materials.
You also need timing.
You need clean signals.
You need a quiet enough biological environment for the embryo to read its own instructions.
Invisible Does Not Mean Inert
One of the most dangerous assumptions in modern health policy is that what we cannot see, taste, smell, or feel must be biologically irrelevant.
But biology itself is built from invisible order.
You cannot see membrane voltage.
You cannot see calcium waves.
You cannot see methyl groups being placed on DNA.
You cannot see mitochondrial redox oscillations.
You cannot see chromatin architecture shifting in three dimensions.
You cannot see ion channels opening and closing with exquisite timing.
Yet all of these invisible processes help determine whether cells divide, migrate, differentiate, repair, inflame, regenerate, or die.
So the invisibility of EMFs is not an argument against biological relevance.
It is the reason the issue is so easy to ignore.
Why Epidemiology Struggles With Meta-Disease
If EMFs create low-fidelity biology, epidemiology will struggle to detect the signal.
Why?
Because the endpoint is not one disease. The endpoint is increased error probability.
The same exposure could be harmless in one person, harmful in another, and catastrophic in a narrow developmental window. The same field could interact differently depending on folate status, B12 status, fever, inflammation, medication exposure, genetic variants, mitochondrial reserve, maternal diabetes, oxidative stress, or the exact waveform and timing of exposure.
Averaging all of that into “exposed” versus “unexposed” groups can erase the very thing we need to measure.
This is why the low-fidelity hypothesis predicts messy data.
Not because the effect is imaginary.
Because the biology is conditional.
The correct unit of analysis is not merely disease incidence. It is signal fidelity inside the biological control system.
The Research We Actually Need
The next generation of EMF research should stop asking only whether a field causes a named disease.
It should ask whether pulsed ELF/RF exposure changes biological fidelity.
That can be tested.
Use blinded, sham-controlled embryo models and cortical organoids. Use precise dosimetry. Use real waveform characterization, not crude frequency labels. Measure calcium timing, membrane voltage, mitochondrial redox state, ROS, chromatin accessibility, methylation, histone acetylation, BET pathway activity, NDEL1 signaling, radial glia differentiation, actomyosin tension, and morphogenetic outcomes.
Test folate and B12 sufficiency as modifiers.
Test CYB5B knockout and rescue.
Test voltage-gated channel perturbation.
Test whether shielding, waveform changes, or optical alternatives reduce biological noise.
Most importantly, define the endpoint correctly:
Does the exposure reduce fidelity?
Not “does every animal get the same disease?”
Not “does every cell die?”
Not “does the tissue heat?”
The question is whether the biological program becomes noisier, less synchronized, more stressed, more epigenetically patched, and less developmentally precise.
That is the science that should have been done decades ago.
The Policy Failure
Public policy has not caught up with this model.
Section 704 of the Telecommunications Act restricts state and local governments from regulating wireless facility placement based on the environmental effects of RF emissions when facilities comply with FCC regulations. Meanwhile, the FDA explains that the Radiation Control provisions originally enacted as the Radiation Control for Health and Safety Act of 1968 apply to electronic products that emit electronic product radiation, including non-ionizing electromagnetic radiation from electronic circuits.
That creates a serious mismatch.
The law recognizes that electronic products can emit non-ionizing radiation relevant to public health.
But the wireless siting regime often prevents communities from acting on health concerns once FCC compliance is asserted.
A low-fidelity biology framework demands a better standard.
Not panic.
Not technophobia.
A better standard.
One that treats timing, waveform, chronicity, developmental windows, pregnancy, infancy, schools, hospitals, and biological signal fidelity as relevant safety parameters.
The Practical Takeaway
The point is not to make people afraid of every signal.
The point is to stop pretending that invisible environmental structure cannot matter.
We should reduce unnecessary near-body pulsed RF exposures during pregnancy and infancy. We should prefer wired connections where practical. We should stop placing transmitters directly against developing bodies. We should design schools, homes, hospitals, and nurseries around lower-noise connectivity. We should accelerate optical and wired alternatives where they make sense. LiFi, fiber, Ethernet, and distance-based design are not anti-technology. They are cleaner technology.
The goal is not to abandon modern communication.
The goal is to stop bathing the most sensitive biological windows in avoidable timing noise.
Angel’s Legacy and the Real Question
For thirty years, the demand has been: “Show me the disease EMFs cause.”
But that was never the right demand.
The right demand is:
Show us whether pulsed, non-native electromagnetic fields degrade the fidelity of biological computation.
If they do, then the downstream disease labels will vary. Neural tube defects, autism features, infertility, immune disorders, cancers, metabolic disease, neurodegeneration, and accelerated aging may not be separate mysteries. Some may be downstream expressions of biological systems forced to operate in lower-fidelity environments.
That is the meta-disease state.
That is the open door.
And that is why this fight matters.
Not because EMFs are the only factor.
Not because every disease has one cause.
But because life depends on high-fidelity signaling, and we have filled the environment with invisible signals without ever properly asking whether biology can still hear itself clearly.
The science now needs to move upstream.
From disease labels to fidelity.
From heat to timing.
From exposure averages to developmental windows.
From denial to measurement.
From “prove disease X” to “measure the noise floor of life.”
That is the low-fidelity biology hypothesis.
And it is time to test it.
