Thirty Original Hypotheses — Incoprea Research

These are hypotheses, not findings. None has been tested here. Several are contrarian; a few are deflationary, arguing that a famous anomaly is ordinary physics wearing a costume. That is deliberate — a hypothesis that cannot lose is not worth writing down.

Scope is physical, chemical, biological and engineered systems only. No social science, no economics, no policy, and no population-level data. The biological entries are about cells, tissues and organisms, not about groups of people.

01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

Matter & Materials 6

Cavitation is an isotope pump, not a furnace

Matter & MaterialsContrarianRisk: MediumSeed: patent + hobbyist replication

Seed: A sonoluminescence reactor patent (US20140290957A1) shows how easily cavitation reactors can be built; the isotope question has been left to hobbyist replication, where claims are rarely separated from chemistry.
Hypothesis: Acoustic cavitation fractionates isotopes by a plasma mechanism, not by evaporation. During collapse the vapor is partly ionized; ions drift in the collapse field with mass-dependent mobility, and recombination preferentially returns light isotopes to the liquid. The fractionation factor per joule should therefore scale with sonoluminescence intensity and electron temperature, and should reverse sign when a high background pressure buffers the collapse and suppresses ionization.
What's new: Sonochemical isotope separation is usually modelled as distillation. This says the plasma chemistry is the pump, which predicts a different scaling law and a pressure-driven sign change.
Decisive test: Sonicate deionized water under argon at 20 kHz and fixed energy; vary static pressure from 1 to 4 bar and dissolved gas; measure δD and δ18O by cavity ring-down spectroscopy alongside sonoluminescence intensity. Prediction: 3–10× the fractionation of thermal distillation per joule, scaling with sonoluminescence, inverting above roughly 2.5 bar.
Kill criterion: No fractionation at fixed cavitation energy, or fractionation that is independent of sonoluminescence intensity and pressure.
Payoff: Turns a fringe claim into a cheap isotope-separation route and gives sonochemical reactor design a real calibration.

Stiction is a rate effect, and the standard measures the wrong number

Matter & MaterialsMethodRisk: LowSeed: reliability standard + patent

Seed: MEMS reliability standards specify quasi-static pull-off tests, while anti-stiction coating patents claim improvements measured at a single speed.
Hypothesis: The pull-off force between a compliant beam and a surface is hysteretic in retraction velocity because the capillary meniscus has a finite drainage time. Adhesion energy follows E(v) = E₀[1 + (kT/E_b) ln(v/v₀)] over four decades of velocity, and the slope depends on humidity, not on coating chemistry. Quasi-static qualification therefore underestimates in-use stiction for devices that separate quickly.
What's new: Capillary adhesion is usually treated as an equilibrium energy. This makes the rate the variable that matters and predicts a universal logarithmic slope set by thermal energy, not by the coating.
Decisive test: Atomic-force-microscope colloidal probe on coated and uncoated silicon, 10 nm/s to 1 mm/s, 20–80% relative humidity; fit the logarithmic law; compare with pull-in voltage on real MEMS beams. Prediction: about kT per decade, independent of coating thickness.
Kill criterion: Pull-off force independent of retraction velocity at fixed humidity.
Payoff: A one-parameter design rule that changes how MEMS are qualified.

Grain boundaries are thermal diodes, and misorientation sets the direction

Matter & MaterialsSynthesisRisk: MediumSeed: patent + dissertation

Seed: Graphene heat-spreader patents assume isotropic, boundary-free conduction; a dissertation on grain-boundary thermal transport reports asymmetric phonon transmission that was never turned into a device.
Hypothesis: Asymmetric tilt boundaries in polycrystalline graphene rectify heat, with a rectification ratio up to 15% at room temperature. The sign is set by whether the boundary is tilt- or twist-dominated, and the effect comes from strain-asymmetric phonon transmission at dislocation cores — not from the graphene lattice itself.
What's new: Thermal rectification in graphene is usually pursued with strain gradients or asymmetric shapes. This makes the grain boundary the device, and gives a sign rule that can be checked by growing films with controlled misorientation.
Decisive test: Suspended CVD graphene with engineered misorientation, Raman thermometry, alternating laser heating on opposite edges; measure the temperature asymmetry. Prediction: rectification above 5% only for boundaries with dislocation cores on one side.
Kill criterion: Less than 2% rectification across all misorientations.
Payoff: Passive thermal diodes for microelectronics and a new knob for phonon engineering.

Hydrogen embrittlement has a percolation threshold, and grain size sets it

Matter & MaterialsSynthesisRisk: MediumSeed: pipeline standard + patent

Seed: Pipeline integrity standards use empirical hydrogen concentration limits, while hydrogen-resistant steel patents optimize strength rather than grain-boundary area.
Hypothesis: Brittle initiation in hydrogen-charged steel is a percolation transition over grain-boundary decohesion sites. The critical hydrogen concentration scales inversely with grain-boundary area, and acoustic-emission avalanches near the threshold show a power-law cluster distribution with an exponent near 1.3. The threshold should be independent of yield strength once grain size is fixed.
What's new: Hydrogen embrittlement is usually framed as a local decohesion or enhanced plasticity mechanism. This turns the onset into a connected-network problem with a measurable critical exponent, and predicts that grain size beats strength as a design variable.
Decisive test: X65 and 316L steels in three grain sizes, electrochemical charging to controlled concentrations, acoustic-emission monitoring, cluster scaling analysis. Prediction: c* ∝ 1/grain-boundary area and an exponent of 1.3 ± 0.15.
Kill criterion: No power-law clustering, or an exponent that changes with microstructure.
Payoff: A microstructure-based charging limit for pipelines to replace empirical thresholds.

The boson peak is a hidden state variable in metallic glasses

Matter & MaterialsContrarianRisk: MediumSeed: relaxation literature + patent

Seed: Metallic-glass annealing studies routinely report samples with identical density but different relaxation behaviour, a residual that fictive temperature does not fully explain.
Hypothesis: Two metallic glasses with identical composition and density but different thermal history have different boson peaks. The peak position and height predict relaxation time and crystallization onset better than fictive temperature, because the glass stores history in medium-range order that density does not capture.
What's new: Fictive temperature is assumed to encode thermal history completely. This claims a second, independent state variable that is directly measurable by inelastic scattering.
Decisive test: Zr–Cu–Al glasses quenched at three rates and annealed below Tg; inelastic X-ray scattering, differential scanning calorimetry, and relaxation measurements on samples matched to 0.1% density. Prediction: boson-peak height separates samples that density cannot.
Kill criterion: Boson peak collapses onto a single function of density.
Payoff: Better annealing schedules and a real handle on metallic-glass aging.

Magic-angle graphene is a strain glass

Matter & MaterialsSynthesisRisk: HighSeed: preprint + patent

Seed: Moiré superlattice preprints show that anisotropic strain reconstructs the superlattice near the magic angle; transport hysteresis in these devices is usually dismissed as charge trapping.
Hypothesis: Near the magic angle, twisted bilayer graphene freezes into a strain-glass phase: moiré domains lock into a distribution of local strains, producing logarithmic aging and a memory dip in resistance after a temperature cycle, analogous to spin glasses but with no magnetic field.
What's new: Moiré systems are treated as clean, tunable crystals. This says the disorder is self-generated and glassy, which would explain irreproducibility across cooldowns and set a limit on device stability.
Decisive test: High-quality twisted bilayer devices; resistance versus time from 10² to 10⁵ s at fixed temperature; temperature cycles with a pause; noise spectroscopy. Prediction: logarithmic aging with a slope that depends on twist-angle disorder, and a memory dip that survives cycling.
Kill criterion: No memory dip beyond trivial drift or charge-trapping hysteresis.
Payoff: Glassy physics in a tunable platform, plus a stability criterion for moiré devices.

Energy & Engineering 6

"Anomalous heat" in metal hydrides is latent heat, and the phase diagram says so

Energy & EngineeringDeflationRisk: LowSeed: declassified review + patent

Seed: The 1989 Department of Energy cold-fusion review and later heat-generation patents report bursts of heat from deuterium-loaded palladium without a consistent chemical accounting.
Hypothesis: The heat bursts are the latent heat of the α→β hydride phase transition, amplified by nanostructuring and non-equilibrium loading. The magnitude is bounded by the pressure–composition–temperature enthalpy; apparent excess comes from poor calorimetry, incomplete loading, and heat retained in the cell.
What's new: It is a boring answer, but it is quantitative: the phase diagram predicts when the bursts occur and how large they can be. That makes the anomaly a calorimetry problem rather than a nuclear one.
Decisive test: Simultaneous calorimetry and four-point resistivity on palladium nanowires while sweeping hydrogen loading. Prediction: every heat burst coincides with the resistivity signature of the α→β transition, total heat stays below the chemical enthalpy, and the effect disappears when alloying suppresses the transition.
Kill criterion: Heat exceeding ten times the chemical enthalpy, or heat correlated with helium-4 or neutron production.
Payoff: Settles a long controversy with a testable chemical explanation and informs hydrogen-storage thermal management.

Osmotic power has an optimum pore size, and it is not the smallest pore

Energy & EngineeringSynthesisRisk: MediumSeed: patent + trade press

Seed: Blue-energy membrane patents chase ever-smaller pores, while published power densities scatter by an order of magnitude at similar pore sizes.
Hypothesis: Power density peaks when the pore radius is about 0.7 of the Debye length. Below that, selectivity is high but conductance collapses; above it, conductance rises but selectivity falls. The optimum therefore shifts with salinity, and pH changes the peak height by tuning surface charge. Much of the literature scatter is unreported pH and Debye length.
What's new: The field optimizes for selectivity alone. This makes the optimum a function of operating salinity, which is a design rule rather than a materials claim.
Decisive test: Membranes with 1–20 nm pores, salinity gradients from 0.01 to 1 M NaCl, pH 3–10; measure power density and confirm the peak tracks the calculated Debye length.
Kill criterion: An optimum that does not move when salinity changes.
Payoff: A concrete design rule for osmotic-power membranes.

Dendrites in solid electrolytes start where the grain boundary conducts electrons

Energy & EngineeringSynthesisRisk: MediumSeed: patent + DOE report

Seed: Solid-state battery patents blame voids and surface flaws for lithium penetration; reports of grain-boundary electronic conductivity in garnets are treated as a nuisance rather than a cause.
Hypothesis: The critical current density for lithium penetration in garnet electrolytes is inversely proportional to the product of grain-boundary electronic conductivity and local stiffness. Dendrites nucleate at grain boundaries, not at pores, and grain boundaries can be ranked by Kelvin-probe force microscopy before cycling.
What's new: Most models treat the electrolyte as an ionic conductor with accidental electronic leakage. This makes that leakage the controlling variable and gives a pre-cycling screening method.
Decisive test: LLZO samples with controlled grain-boundary chemistry; Kelvin-probe force microscopy maps; operando X-ray tomography during plating; measure critical current density per boundary type.
Kill criterion: Dendrites nucleate at pores regardless of grain-boundary electronic conductivity.
Payoff: A fast screening method for solid electrolytes.

Rain erosion obeys a chemistry-independent velocity exponent

Energy & EngineeringSynthesisRisk: MediumSeed: IEC standard + trade conference

Seed: Wind-turbine blade qualification uses IEC rain-erosion test sequences that compare coatings at a few velocities, without a shared scaling law.
Hypothesis: Leading-edge erosion mass loss scales as velocity to the fifth power for every coating, because the mechanism is fatigue crack growth in the coating and the exponent is inherited from the fatigue law. Coatings differ in threshold velocity and toughness, not in exponent. Normalizing by threshold should collapse all coatings onto one curve.
What's new: Erosion is usually treated as a materials-specific empirical curve. This predicts a universal exponent and a collapse, which would cut the number of tests needed.
Decisive test: Whirling-arm and single-impact jet tests on five coatings from 70 to 140 m/s; measure mass loss and pit depth; fit exponents and test the collapse.
Kill criterion: Exponent varies by more than ±20% across chemistries.
Payoff: A universal lifetime model and far fewer qualification tests.

Additive manufacturing stores a thermal memory that fatigue can read

Energy & EngineeringContrarianRisk: HighSeed: trade manual + patent

Seed: Laser powder-bed trade manuals treat scan strategy as a distortion-control knob; fatigue specifications assume porosity and surface finish are the only defect variables.
Hypothesis: Fatigue life depends on the scan-order path integral of peak temperature, not only on porosity. Two parts with identical geometry and identical porosity can differ by a factor of two in fatigue life, and the difference survives hot isostatic pressing and machining because it is encoded in dislocation substructure.
What's new: The industry equates a good part with a low-porosity part. This says the thermal path is an independent design variable with a mechanical signature.
Decisive test: 316L samples built with two scan strategies, matched porosity by computed tomography, HIPed and machined identically, then fatigue-tested; predict life from thermal simulation before testing.
Kill criterion: Fatigue life collapses onto porosity and surface finish alone.
Payoff: Scan strategy becomes a fatigue design variable, not just a distortion fix.

Supercapacitor fade is ion pairing, and it is partly reversible

Energy & EngineeringContrarianRisk: MediumSeed: patent + trade journal

Seed: Electrolyte-additive patents claim cycle-life improvements without agreeing on whether the electrode or the electrolyte is failing.
Hypothesis: Capacitance fade is dominated by ion-pair clustering in the electrolyte, not by electrode surface loss. Fade follows a stretched exponential with an exponent near 0.4 across chemistries, and a controlled voltage hold can de-cluster the electrolyte and recover 10–30% of the lost capacitance.
What's new: Aging is usually attributed to electrode degradation. This makes the electrolyte the reversible part and predicts a shared exponent across unrelated devices.
Decisive test: Cycle commercial cells, track capacitance and resistance, use nuclear magnetic resonance and conductivity to quantify ion pairing, then apply recovery holds; compare with post-mortem surface-area measurements.
Kill criterion: Fade correlates with surface-area loss and no recovery is observed.
Payoff: A maintenance protocol and a clearer target for electrolyte design.

Earth & Planetary 6

Methane seeps are paced by the cap, not the source

Earth & PlanetarySynthesisRisk: MediumSeed: IODP report + journal

Seed: Hydrate-controlled seepage studies show episodic methane release, but the source is usually assumed to be the clock.
Hypothesis: Seep periodicity is set by permeability hysteresis in the hydrate cap. As hydrate forms, the cap self-seals; pressure builds; the cap fails; the cycle repeats. The period scales with cap thickness squared, and seeps should cluster in bands of cap thickness rather than in bands of source strength.
What's new: It makes the cap the oscillator. That predicts a geometric scaling law and a spatial pattern that can be tested against existing seep chronologies.
Decisive test: IODP cores and pore-pressure monitoring at three seeps with different cap thicknesses; laboratory flow tests on hydrate-bearing sediment; compile published seep chronologies and test the thickness-squared relation.
Kill criterion: Period independent of cap thickness.
Payoff: A predictive rule for seafloor stability and carbon flux.

Fulgurites are continental paleoatmosphere recorders

Earth & PlanetaryMethodRisk: MediumSeed: GSA abstract + museum collection

Seed: Fluid-inclusion analyses of fulgurite glasses show trapped gases, but the material has been treated as a curiosity rather than an archive.
Hypothesis: Lightning implants contemporary atmosphere into quartz glass. Noble-gas ratios in dated fulgurites record the atmosphere at formation, with a correctable mass-dependent fractionation, giving a continental record independent of ice cores.
What's new: Paleoatmosphere work depends on ice and sediments. This proposes a dated, continent-wide archive that exists in museum drawers.
Decisive test: Step-heating noble-gas mass spectrometry on fulgurites dated by luminescence and radiocarbon from three continents; compare argon, nitrogen, and xenon ratios with ice cores over the same intervals; quantify and correct fractionation using laboratory lightning analogues.
Kill criterion: Gases are dominated by parent-rock volatiles or metamorphic overprinting.
Payoff: A new paleoclimate archive, especially for intervals older than the ice cores.

Martian perchlorate is made by dust, not sunlight

Earth & PlanetaryContrarianRisk: MediumSeed: NASA report + discharge paper

Seed: Mars perchlorate is usually attributed to ultraviolet photochemistry, but discharge experiments in simulated dust devils produce perchlorate at rates that were never compared directly.
Hypothesis: Triboelectric discharges in dust devils and storms produce most of the perchlorate from chloride-bearing dust, and the rate is controlled by the charge-to-mass ratio of the grains rather than by total dust flux. Coarse, chloride-rich grains make more perchlorate per unit mass than fine dust, so the global budget tracks grain-size distribution, and ultraviolet photochemistry is secondary.
What's new: Discharge chemistry has been proposed before; this makes grain size the controlling variable and predicts a spatial pattern that the ultraviolet model does not.
Decisive test: Mars chamber with size-sorted, chloride-doped dust, controlled agitation, discharge counting, and perchlorate measurement; compare predicted production with Curiosity and Perseverance surface data and with mapped grain-size distributions.
Kill criterion: No perchlorate production above the ultraviolet-only control.
Payoff: Explains the spatial pattern of perchlorate and informs both toxicity and resource use.

Nanophase iron is the microwave handle on the Moon

Earth & PlanetarySynthesisRisk: LowSeed: patent + agency report

Seed: Lunar construction patents claim microwave sintering of regolith, but heating rates vary widely between simulants and actual regolith.
Hypothesis: Microwave heating of lunar regolith is dominated by nanophase metallic iron produced by space weathering. Heating rate scales with nanophase iron volume fraction, and prior reduction can tune it. This explains why some regolith heats faster than its simulant.
What's new: The field treats microwave absorption as a bulk dielectric property. This identifies a specific, controllable nanophase as the susceptor.
Decisive test: Regolith simulants with controlled nanophase iron fractions, dielectric measurements from 1 to 30 GHz, microwave heating curves, and comparison with Apollo sample data.
Kill criterion: Heating rate independent of nanophase iron content.
Payoff: A design rule for lunar construction and in-situ resource use.

Earthquake lights are p-hole recombination, and quartz content predicts them

Earth & PlanetarySynthesisRisk: HighSeed: USGS report + historical catalog

Seed: Historical earthquake-light catalogs are rich but statistically messy; the positive-hole theory of stressed rock has laboratory support that has never been connected to the light catalog.
Hypothesis: Stressed quartz generates positive holes that migrate to the surface and recombine, emitting light. Earthquake-light occurrence correlates with quartz content and stress rate, not with fault geometry or distance from the epicenter alone.
What's new: It makes a specific material prediction — quartz content — rather than a generic "stress produces light" claim, and it can be tested in the laboratory before the field.
Decisive test: Deform quartz-rich and quartz-poor rocks in a press with a photomultiplier and electrometer; then compare a curated earthquake-light catalog against regional geology and stress-rate estimates.
Kill criterion: No light emission from stressed quartz at relevant stress and strain rates.
Payoff: A physical mechanism for a persistent anomaly and a possible strain indicator.

Hessdalen is a dusty plasma sustained by rock piezoelectricity

Earth & PlanetaryContrarianRisk: HighSeed: field-report archive

Seed: The Hessdalen valley in Norway has produced luminous phenomena since at least the 1930s; the field archive is large but the explanations remain fragmented.
Hypothesis: The lights are a dusty plasma maintained by field emission from piezoelectric charge separation in the valley's rocks, modulated by humidity. The spectrum should show characteristic plasma bands plus a dust continuum, and activity should correlate with humidity and rock stress.
What's new: It replaces "ball lightning" and "aircraft" with a specific, local, material mechanism that makes a spectral prediction.
Decisive test: Field spectroscopy with a slitless spectrograph, simultaneous rock piezoelectric measurements, humidity and electric-field logging; compare spectra with known atmospheric plasmas and aircraft lights.
Kill criterion: Spectra match known mundane sources, or no piezoelectric signal is found in the valley rocks.
Payoff: Resolves a decades-old anomaly, or kills it cleanly.

Life & Biology 6

Planaria write their target morphology into gap junctions, and it can be rewritten

Life & BiologySynthesisRisk: MediumSeed: bioelectricity preprint + patent

Seed: Bioelectricity work shows that gap junctions carry patterning signals in regenerating tissues; the "target morphology" is inferred but its storage and erasure are not localized.
Hypothesis: Target morphology is stored in the gap-junction network's electrical state. A transient gap-junction block during a specific regeneration window rewrites that state, producing a heritable change in regenerate shape. The sensitive window is set by the voltage edge at the cut, not by the clock alone.
What's new: It treats the target morphology as addressable memory with a defined write window, which makes it an experiment rather than a metaphor.
Decisive test: Cut planaria, expose to octanol in windows from 0 to 48 hours after cutting, score regenerate shapes over three generations, and correlate with voltage-sensitive dye maps.
Kill criterion: No persistent shape change beyond normal variation.
Payoff: Bioelectric control of regeneration and a test bed for morphological memory.

Mycelium computes best at the edge of percolation

Life & BiologySynthesisRisk: MediumSeed: mycology preprint + patent

Seed: Fungal electrical-signaling papers report spike-like activity, and fungal-material patents exploit mycelium networks, but the network has not been tested as a computational substrate with a density knob.
Hypothesis: Mycelial networks act as fading-memory reservoirs. Their computational capacity peaks at a specific hyphal density near the percolation threshold, because that is where signal paths are long but not fragmented, and capacity can be predicted by a single-timescale reservoir model.
What's new: It makes density a design parameter and predicts a peak rather than "more network is better."
Decisive test: Grow networks at controlled densities on patterned substrates, record electrical responses to stimulus histories, and benchmark against reservoir models; measure memory capacity and nonlinearity.
Kill criterion: No fading memory, or capacity independent of density.
Payoff: Fungal materials as sensors and computers, with a concrete design target.

Bacterial nanowires are proton wires, not electron wires

Life & BiologyContrarianRisk: HighSeed: dissertation + microbiology preprint

Seed: A dissertation on microbial cytochrome nanowires documents electron transport, while other measurements report a strong humidity dependence — a signature that has not been reconciled.
Hypothesis: Geobacter nanowires conduct protons along surface carboxyl groups, not electrons. The apparent electronic conductivity is a contact-geometry artifact. Conductivity should show a deuterium isotope effect, a strong humidity dependence, and pH titration behavior consistent with proton hopping.
What's new: It challenges the bioelectronics premise that these filaments are biological wires and offers three cheap falsifiers.
Decisive test: Measure conductivity in H₂O versus D₂O vapor, across humidity from 10% to 90%, and as a function of pH; control contact geometry and tip force.
Kill criterion: Conductivity insensitive to both deuterium substitution and humidity.
Payoff: Rewrites a bioelectronics premise and points to protonic materials.

Anhydrobiosis is decided in the mitochondria

Life & BiologyContrarianRisk: MediumSeed: astrobiology report + patent

Seed: Desiccation-tolerance work focuses on cytosolic trehalose glass, but survival varies widely at similar trehalose loads.
Hypothesis: Survival after desiccation is set by vitrification of the mitochondrial matrix, not by bulk cytosolic trehalose. Matrix viscosity predicts survival, and cristae collapse is the lethal event. Trehalose is necessary but not sufficient.
What's new: It moves the critical glass from the cytosol to an organelle, which is directly testable with targeted probes.
Decisive test: Express FRET viscosity probes targeted to mitochondria and cytosol; desiccate cells; correlate matrix viscosity with survival and with cristae morphology by electron microscopy.
Kill criterion: Survival correlates only with bulk trehalose concentration.
Payoff: Better preservation of cells and vaccines, and a sharper target for astrobiology.

Cryptochrome needs the right membrane, not just the right protein

Life & BiologySynthesisRisk: MediumSeed: magnetoreception preprint + patent

Seed: In vitro cryptochrome experiments give inconsistent magnetic sensitivity, while membrane composition is rarely controlled.
Hypothesis: The radical-pair lifetime that underlies magnetoreception is set by membrane viscosity and lipid order, not by protein structure alone. Magnetic sensitivity peaks at a specific cholesterol content, and protein-only assays cannot reproduce it.
What's new: It makes the lipid environment a functional component of the sensor and explains contradictory reconstitution results.
Decisive test: Reconstitute cryptochrome in liposomes of varying viscosity and cholesterol content; measure radical-pair lifetime by transient absorption and magnetic-field sensitivity by flash photolysis.
Kill criterion: Radical-pair lifetime independent of membrane viscosity.
Payoff: A design rule for magnetic biosensors and a resolution of the in vitro disagreement.

Pressure folds the dark proteome

Life & BiologySpeculativeRisk: HighSeed: piezophile literature + patent

Seed: Piezophile enzymes are known to be pressure-adapted, but the broader disordered proteome has not been screened for pressure-induced folding.
Hypothesis: A subset of intrinsically disordered proteins adopts stable, functional folds at 1–3 kbar. These conditional folds are cryptic at atmospheric pressure and may explain piezophile function and pressure-stable enzyme activity.
What's new: Disorder is usually treated as a pressure-independent property. This proposes a hidden fold space that is accessible only under pressure.
Decisive test: Small-angle X-ray scattering and nuclear magnetic resonance at 1–3 kbar on 50 disordered proteins; pair with activity assays under pressure.
Kill criterion: No disorder-to-order transition in the screened set.
Payoff: A hidden fold space and a route to pressure-stable biocatalysts.

Computation & Information 6

The best exceptional-point sensor is not at the exceptional point

Computation & InformationContrarianRisk: MediumSeed: arXiv preprint

Seed: A 2025 analysis shows that noise and nonlinearity remove the divergences that exceptional points were supposed to provide, leaving a finite noise floor.
Hypothesis: Because noise shifts the exceptional point and the noise floor is finite, signal-to-noise ratio peaks at a detuning proportional to the square root of the noise strength, not at zero detuning. The optimum can be predicted from the fluctuation Hamiltonian and should be visible as a broad maximum rather than a divergence.
What's new: It accepts the 2025 result and turns it into a design rule: operate off the exceptional point by a calculable amount.
Decisive test: Coupled saturable resonators with tunable noise; measure signal-to-noise ratio versus detuning across four decades of noise power; compare with the predicted square-root optimum.
Kill criterion: Signal-to-noise ratio is monotonic in detuning, with no interior maximum.
Payoff: Rescues exceptional-point sensing with a concrete operating point.

Analog PDE solvers only win on ill-conditioned problems

Computation & InformationSynthesisRisk: MediumSeed: patent + conference

Seed: Analog in-memory computing patents claim order-of-magnitude speedups for differential equations, but benchmarks use well-conditioned toy problems.
Hypothesis: The analog advantage scales with the condition number of the discretized operator. Below a condition number near 10³, conversion and calibration overhead dominates; above it, analog wins by orders of magnitude because the iteration count of the digital solver grows with conditioning.
What's new: It predicts when analog wins, rather than asserting that it does.
Decisive test: Memristor or phase-change crossbars solving Poisson and heat equations with condition numbers from 10 to 10⁵; benchmark time-to-solution and energy against a digital solver at matched accuracy.
Kill criterion: No dependence of speedup on condition number.
Payoff: A decision rule for when to build analog hardware.

SRAM entropy is predictable from transistor mismatch

Computation & InformationMethodRisk: LowSeed: hardware-security standard + patent

Seed: Physical unclonable function patents rely on measured entropy, while SRAM power-up bias is usually corrected with whitening rather than predicted.
Hypothesis: Power-up bias is predicted by Pelgrom mismatch coefficients. A bias-corrected extractor can pass the NIST SP 800-90B tests without whitening, and entropy per bit is a computable function of threshold-voltage mismatch.
What's new: It replaces measurement-and-whiten with a priori prediction from a standard mismatch model, which changes how entropy is certified.
Decisive test: One hundred chips from two process nodes; characterize mismatch; collect power-up patterns; run the NIST suite on bias-corrected output; compare predicted and measured min-entropy.
Kill criterion: Bias uncorrelated with mismatch coefficients.
Payoff: Cheaper true random number generation and defensible entropy estimates.

DNA storage can purify itself with mismatch-binding proteins

Computation & InformationMethodRisk: MediumSeed: synthetic-biology patent + preprint

Seed: DNA storage error correction is almost entirely computational; enzymatic error removal before sequencing is rarely used.
Hypothesis: Error-containing strands can be removed before sequencing: hybridize the library to a perfect reference complement, then capture the resulting mismatches on immobilized MutS. Enrichment scales with the number of mismatch sites and wash stringency, giving a hundredfold error reduction at twofold redundancy.
What's new: It moves error correction from the decoder to the wet lab, where it can be cheap and parallel.
Decisive test: Synthesize oligos with known error rates, pass them over MutS columns under varied salt and temperature, sequence before and after, and measure enrichment.
Kill criterion: No enrichment above input error rate.
Payoff: Lower sequencing cost and longer effective storage lifetimes.

Time-reversal focusing gain is set by the waveguide invariant

Computation & InformationSynthesisRisk: MediumSeed: navy report + patent

Seed: Time-reversal acoustics demonstrations report focusing gains that vary widely between shallow-water sites, usually attributed to array size and bottom loss.
Hypothesis: In shallow water, time-reversal focusing gain is proportional to the waveguide invariant, not to array size alone. The optimal frequency is set by mode attenuation, and the site-to-site scatter should collapse onto the invariant.
What's new: It gives a single environmental parameter that predicts gain, replacing trial-and-error array design.
Decisive test: At-sea time-reversal experiments at three shallow-water sites with different invariants; measure gain and invariant independently; test the proportional relation across frequencies.
Kill criterion: Gain independent of the waveguide invariant.
Payoff: Predictive design for sonar and underwater communication.

Conductance drift is a forgetting factor you can use

Computation & InformationContrarianRisk: MediumSeed: neuromorphic patent + trade press

Seed: Neuromorphic engineering treats conductance drift in analog memory as the central reliability problem and compensates it away.
Hypothesis: Drift implements a Bayesian forgetting factor. Reservoirs with finite drift have a longer effective memory horizon and better performance on non-stationary tasks, and performance peaks at a finite drift rate rather than at zero drift.
What's new: It turns the main defect of analog memory into a computational resource with an optimum.
Decisive test: Reservoir computing on a non-stationary benchmark with controlled drift rates spanning two decades; measure task performance and memory horizon; fit a Bayesian forgetting model.
Kill criterion: Performance decreases monotonically as drift increases.
Payoff: Drift becomes a design parameter rather than a defect to eliminate.