Eskridge-Amy: Arbiter (STEP 3)

Date: 2026-10-07. Inputs read in full: claims.md, physicist.md, mathematician.md, contrarian.md, facts.json (via the reviews), sites.md and the Investigator's notes as summarised in the reviews; I also re-read the NASA TM Table 1 (extracted text p.25) and the deep-read notes on Appendix B/C. Tags: [read] project document, [web] URL, [memory] with confidence, [calc] my own calculation (Python, 2026-10-07).

Source gap (repeated from claims.md)

Amy Eskridge published no paper, preprint, patent, equation or data. Her own material is a slide deck (HAL5, 6 Dec 2018), a machine-captioned lecture, a 3 h 15 min machine-captioned interview (about May 2020, no speaker labels) and the Institute website. None contains an equation, data table or protocol of hers. Everything about her physics rests on her unsupported statements. The only technical documents are by others: her father's NASA TM-20205010911 (she is not named) and the Tajmar-group paper (arXiv 2402.15529). Not obtained: any MSFC report of the 2015 pulsed-superconductor test, Alzofon's 1981 AIAA paper 81-1608, the interview audio, the Sept 2020 cease-and-desist text, the Dembovskis and Prutchi documents, the Shellenberger testimony. Where a verdict below rests on [web]/[memory] and not on her documents, it says so.

Interested parties (their descriptions are reported as theirs): Amy Eskridge and the Institute (sought investors); Richard Eskridge (father, co-author of the TM); Quantum Machines LLC (funded POAMS work; holds the "V5" data); Mark Sokol, Falcon Space and the APEC conference (Alzofon/DNP test; his own conference); Tim Ventura's APEC articles; The Sentinel Network (suppression thesis); Podkletnov and Poher (authors of their own claims).

Where I recomputed because the reviewers differ

1. Thermal polarisation of Bi-209 at 0.2 T, 300 K. Physicist: 1.8e-6 (and a shortfall of 200-450x). Mathematician: 4e-7 (shortfall about 2000x). Contrarian: 4e-7 (shortfall about 1000x). [calc] Bi-209: I = 9/2, mu = 4.11 nuclear magnetons, so gamma = 4.37e7 rad/s/T. x = mu B/kT = 1.00e-6. The exact Brillouin sum gives /I = 4.08e-7. The physicist's 1.8e-6 is (I+1)x/3, which is in units where the full-alignment value is I, so it is a factor I = 4.5 too large when compared with the "fraction fully aligned" the TM's factor 1.69 assumes. The mathematician and contrarian are right. Required aligned fraction for 0.3 g on 211 g: 0.142%/1.69 = 8.4e-4 (co-spin) or 4.3e-4 (using 3.29). Shortfall: 8.4e-4 / 4.08e-7 = about 2.1e3 (and 1.1e3 using 3.29). Predicted weight change by the TM's own force law with thermal alignment: 4.08e-7 x 1.69 x 211 g = 1.5e-4 g, against 0.3 g reported and a 0.01 g scale resolution. Other conditions [calc]: 9 T, 300 K gives 1.8e-5; 0.2 T, 5 K gives 2.5e-5; 9 T, 5 K gives 1.1e-3 (only a cryogenic high-field magnet would reach the needed level, and the TM describes a 0.2 T room-temperature rotor). Final figure: about 2000x too small (1000x at the most generous reading), not 500x.
2. Barnett (rotation) alignment. Effective field omega/gamma: 7.2 microtesla at 3000 rpm (contrarian) and 12 microtesla at 5000 rpm (physicist) are both right; polarisation about 1e-11, negligible against the 0.2 T magnet. The contrarian's "1e-10" is a rounding up; no verdict depends on it.
3. Aluminium DNP (Alzofon). [calc] Al-27 thermal polarisation at 0.33 T, 300 K: 6.8e-7 (mathematician: about 7e-7, confirmed). Full DNP enhancement gamma_e/gamma_Al is about 2500, giving about 1.7e-3 at most. Alzofon's reported 0.22% change would then need about 100% weight change per polarised nucleus. As the mathematician said, no theory in the folder derives that.
4. TM Table 1 (gyroscope arm force). [SUPERSEDED: this item is wrong; see QE revision A1. Table 1 IS reproduced by eq. 28a once the sign convention is applied. My original text is kept unchanged below so the error stays visible.] Mathematician: Table 1 is reproduced by eq. 28a and satisfies (F_N + Df_co)^2 + (F_N + Df_anti)^2 = 2 F_N^2 exactly, and the deep read's contrary statement is a normalisation error. Physicist and deep read: not reproduced above 10 Hz. [calc] I took the table's co-spin and anti-spin forces, subtracted the zero-spin force 116.32 N to get Df, and tested the identity with F_N = 89.85 N (the disc-only force the mathematician names): sum of squares over F_N^2 = 2.0001, 2.0023, 2.0119, 2.0378, 2.0941, 2.2020 for 10 to 60 Hz. With the 116.32 N normalisation (the deep read's): 2.0001, 2.0016, 2.0081, 2.0259, 2.0644, 2.1383. Neither normalisation gives 2.0000 above 10 Hz. The mathematician's correction is not confirmed; the deep read and physicist are right that the table is not an exact 28a output at higher spin. Also, solving the table's co- and anti-spin entries for the "spin term" gives different values for the two branches (e.g. 32.1 N vs 50.2 N at 60 Hz), so the Appendix B sheet must use something beyond the printed eq. 28a, or the branches are treated unequally. The deep read says the sheet reproduces the table [read: deep/NASA TM, Appendix B notes]. Status: the table is reproducible from the sheet but not from the printed formula; the discrepancy is unexplained in the TM. No verdict turns on it, because no experiment reached the table's regime (the V1 gyroscope spun down from bearing friction, and results were "meaningful measurements ... could not be made" [read: TM p.25]).
5. Magnetic field of a 20 kA pulse at 5 cm [calc]: B = mu0 I/(2 pi r) = 0.08 T (contrarian, confirmed). Mass needed for a 0.05% and 2% change in g at 1 m [calc]: 7.3e7 kg and 2.9e9 kg (physicist, confirmed). LN2 flash energy: 200 J (100 uF at 2 kV, an assumed capacitance) vaporises about 1.2 mL of liquid nitrogen (physicist's figure; arithmetic consistent: latent heat about 160 kJ per litre [memory: high]).

Corrections to the claim list that all later steps should carry

• "Six forces" attribution. [CORRECTED by QE revision A5: the speaker is most probably Amy.] The speaker at interview [00:20:20-00:20:50] lists "eight forces of nature" and says "my group has a mathematical equation". The same passage has a speaker referring to "her dad" in the third person (mathematician) and one who "pestered him ... in Jersey" (contrarian, a passage near the Tajmar story). The transcripts have no speaker labels, so the speaker may be Mark Sokol or Jeremy Rys. Attribution to Amy is uncertain. The same applies to the "German Air Force" remark (claim 4) [corrected by A5: she says it herself at 01:05:49].
• "Alzofon" in the May 2020 plan is an inference: the caption reads "alzheimer's experiment ... algebra high power here in jersey" [read: interview 02:10:02]. The link to Alzofon rests on APEC [web: https://www.altpropulsion.com/mark-sokol-anti-gravity-with-present-technology/, interested party] and is plausible.
• Dresden test of Alzofon. It used plain NMR pulses at 9 T and about 5 K on 5 mm spheres, not microwave DNP at 0.33 T, room temperature, aluminium powder with iron inclusions. It did not test Amy's planned experiment and does not mention her. No-specimen controls reproduced the deflection and tracked temperature; the refractive-index explanation accounts for about 70% of the signal and lacks the path length and pressure needed to close it [read: deep read of arXiv 2402.15529]. claims.md claim 3 should say "NMR variant" and keep this.
• Correct reading of the TM bismuth passage [read: TM p.33, p.35-36]: 0.01 g OHAUS scale, powered by wires, 0.3 g loss then 0.3 g gain after disassembly and reassembly; the first loss "did not inspire much excitement because this weight loss could be due to thermal buoyancy or aerodynamic forces"; the effect "lessened each time", disappeared after several tests, "returned with equal force" with new magnets, and was later suspected to be a magnet phenomenon; attendees Richard Eskridge, Nelson and Chris Milam (CEO of the funder); the TM calls the data "rudimentary and preliminary"; V5 data remain with Quantum Machines. Include these in claims.md claim 3.
• "500 g" is ambiguous in a machine transcript (g-force or grams) [read: lecture 00:49:38]; as g-force it is a shock-class reading.
• The Poher match to the lecture's "Boyer/Pod Koyer" device is an inference; Poher's device is independent of Podkletnov's, not a replication of it [web: https://arxiv.org/pdf/2206.07574].
• Mansfield Amendment. [web] Section 203 of the FY1970 Defense Procurement Authorization Act (Public Law 91-121, 1969) limited DoD research funds to projects with "a direct and apparent relationship to a specific military function or operation" (search result summary, e.g. https://www.gao.gov/products/D12080 listed). A 1973 Mansfield amendment to the military procurement authorization exists as a Senate vote (https://voteview.com/rollcall/RS0930410) but I did not retrieve what it changed. The wording does not mention gravity; her "illegal to investigate anti-gravity" is her interpretation (the lecture says "there's different interpretations" [read: lecture 00:11:38]).
• Sentinel misexpands POAMS; the TM title is "Pope-Osborne Angular Momentum Synthesis".
• The Investigator's "Li 1997 did not reproduce Podkletnov" should read: a static levitation test, acceleration change under 2 parts in 10^8 (Physica C 281, 260-267 [web: https://ntrs.nasa.gov/citations/19990039542]); it bears on Podkletnov only indirectly.

Web verification done (budget 8; 7 used)

Corrected claim list with final verdicts

Claim 1. Gravity modification is real, was found, then suppressed or lost

Agreement. All three: no mathematics of hers to check; the slides list devices of very different quality and she herself marks Brown, Wallace, Hutchison and Searl as "unsubstantiated", doubts the 20% Morningstar rumour and says Laithwaite's effect is explained by Newton's second law [read: slides pp.10, 16, 18; lecture 00:09:34, 00:31:10]. Poher-type impulses with double shielding: artifact. The suppression thesis is not testable as framed and each case has an ordinary explanation that fits at least as well. Disagreement. Physicist: Podkletnov "Tested, not supported" for every reduced version. Contrarian: "Never properly tested" for the rotating-disc claim, because nobody built the same device (static test; Hathaway at 400-800 rpm, no RF drive). Which is stronger. The contrarian's, with the physicist's facts. The nulls (Li 1997, Woods 2001, Hathaway 2003) are real and published with methods [web], but none copies the 5,000 rpm, RF-driven, cooled, composite disc. The claimant never supplied a specification, video or raw data that would let anyone copy it (physicist; confirmed by the Gallerati review). Equally, a 0.05-2% effect would be enormous: 7e7 to 3e9 kg of ordinary gravitating mass at 1 m [calc], and gravimeters resolve far smaller changes, so the claim's size makes mundane causes (buoyancy, convection above liquid nitrogen, magnetic force on the test mass) the first suspects; she raised convection herself [read: lecture 00:42:27]. Suppression evidence kept in the record. Podkletnov's dispute with Tampere University and the withdrawn paper [web: https://arxiv.org/abs/cond-mat/9701074 via the contrarian's search]; Ning Li: nothing published after 2002, AC Gravity licence renewed through 2018, injured 2014, died 27 July 2021 [web: Huntsville Business Journal, https://huntsvillebusinessjournal.com/news/2023/07/30/solving-the-mystery-of-huntsvilles-brilliant-scientist-disappearing/; Sentinel (interested, suppression framing)]; Robertson (an Institute adviser) said NASA's rotating-disc replication was "never completed" (via APEC, interested) [web]. Her slide on Ning Li's disappearance cites one anonymous Reddit post and she says "I don't know what happened" [read: slide p.25] [tag corrected by A5: the words "I don't know what happened" are in the lecture at 00:18:16; slide p.25 is the Reddit screenshot]. What would settle it. A funded, faithful Podkletnov replication (composite disc, about 5,000 rpm, RF field, cooled, calibrated gravimeter, magnetic and convective controls, raw data public). Robertson's 2025 replication (per APEC) is the live candidate and is unreported. Final verdict (split). - Podkletnov rotating-disc weight loss, original conditions: Never properly tested (not dismissed without testing: partial versions were tested and were null). - Reduced or modified versions (static test, slow rotation, no RF): Tested, not supported. - Poher/impulse discharge device: Tested, not supported (EM artifact and acoustic noise under shielding; Poher's own reports are the only positives). - Hayasaka-type gyroscope weight change: Tested, not supported (1990 nulls). - Brown, Wallace, Hutchison, Searl, Morningstar: Never properly tested (no data or protocol; she herself doubts most). - "Found, then suppressed or lost" as a general claim: Not testable as stated; Ning Li is an open historical case (see Absence of evidence).

Claim 2. Her father's unpublished 2015 MSFC test of a Poher-type pulsed YBCO device

Agreement. All three: this is documented by her statement, one unnamed first-hand voice in the lecture Q&A, and NASA-published author biographies saying Eskridge and Schoenfeld "researched the anomalous gravitational effects of electrically pulsed superconductors" (a statement that such work occurred; no date, data or result) [read: lecture 00:29:08-00:30:40, 00:49:38-00:52:44; TM pp.37-38]. The first-hand voice reports a null on Poher's signals under a double Faraday cage and warns "getting a signal is never a problem ... it's a question of getting the right signal" [read: lecture 00:51:12]. The dewar failure, thrown magnet and shock reading have ordinary candidates: flash boiling (200 J boils about 1.2 mL of LN2 [calc]), a transient field of about 0.08 T at 5 cm from 20 kA [calc], and steam shock. Disagreement. None on the verdict; they differ in how to describe the silence. The contrarian notes a possible conflation with the 1 July 2015 Quantum Machines Space Act Agreement (the only dated 2015 MSFC agreement in the record [read: TM p.5]). The speaker "possibly her father" is unconfirmed. Stronger argument. Both: it was not rejected, it was not published, and the speaker's own account is less favourable to a gravity reading than her slide ("very similar results" [read: lecture 00:30:10]). What would settle it. The MSFC test notes, accelerometer traces and the 1,000,000 frames-per-second film; a Space Act Agreement with the French partner if one exists; a control discharge into the same dewar with a warm or non-superconducting disc; Richard Eskridge, who is alive and speaks to the press [web: NewsNation], could say who the speaker was. Final verdict: Never properly tested (a real but informal and unreported test; no controls, protocol or data on record; the reported effects are not shown to be non-electromagnetic, and the one core null is stated by the account itself). Not dismissed by anyone.

Claim 3. Aligning nuclear spins can change weight; this is the next thing to test

Agreement. The TM algebra (eqs 11-28a) is correct and equals the POAMS postulate Gbar/G = v/v_o restated; the physics is in the postulates (natural orbit, spin energy added to orbital kinetic energy with sign set by spin direction although kinetic energy is a scalar, extension to any body-centred force, ad hoc reversed root 28b). The bismuth experiment was uncontrolled and preliminary with a fading, reassembly-dependent effect. The Dresden group's NMR test was null at low pulse counts with a heating artifact at high. Disagreement 1 (verdict on the theory). Physicist: Refuted by calculation (about 500x). Contrarian: Refuted by calculation at the TM's own numbers, pending recheck. Mathematician: Internally inconsistent in application; gyroscope form refuted by Euler's first law. My recomputation (item 1 above) corrects the physicist's figure to about 2000x and confirms the contrarian's and mathematician's. The calculation is conditional: it assumes thermal polarisation is the only alignment mechanism, which the TM does not quantify beyond "very weak method of excitation" [read: TM p.33]. Barnett alignment is negligible [calc]. Within the TM's own framework, then, the reported 0.3 g is about 2000 times what its force law predicts for the available alignment. That is a refutation by arithmetic of the claim "the data match the theory", not of every conceivable spin-weight coupling. Disagreement 2 (internal consistency). Three results for one bismuth nucleus (2.69, 1.97, 3.19 times weight, depending on route; mathematician), the sign convention flips between the gravity example and the gyroscope table, the r = r_o assumption is applied at the Earth's surface where r is about 290 times r_o, the main-text Hayasaka factor (0.999999962) differs from Appendix A (0.999999946), and Appendix C normalises by mg and not by its own F_N. All agreed except that the physicist did not list the three-way result; I accept the mathematician's table. The TM's reason for suspending disbelief ("surface velocity can easily exceed the speed of light") is wrong for bismuth: the classical 7.12 fm sphere at 7.46e19 rad/s gives 5.3e5 m/s, 0.18% of c [calc by deep reader and mathematician]. Table 1: [CORRECTED by A1: reproduced exactly by eq. 28a with the sign convention Fmeas = Fgyro - DF; the earlier statement here that it is unexplained was wrong]. It remains unmeasured. Disagreement 3 (Euler's first law). Only the mathematician raised it: for a rigid body, total external force = total mass times centre-of-mass acceleration, whatever the spin, so a gyroscope's arm force at fixed orbit should not change with spin rate. This is correct mechanics. It means POAMS's gyroscope prediction (15-19% at 60 Hz) is a denial of accepted mechanics, so it needs a new force, not just a reformulation. It is not a closed refutation of the claim, because the claim is exactly that such a force exists; the experiment decides, and the gyroscope experiment never reached a measurable regime (bearing friction, 50 N load cell) [read: TM p.25]. I therefore record the gyroscope form as Never properly tested, with a statement that it contradicts established rigid-body mechanics and that Hayasaka-type nulls (Quinn and Picard, Faller, Nitschke and Wilmarth [web]) bear on it indirectly. Disagreement 4 (Alzofon DNP). Physicist: Tested, not supported. Contrarian and mathematician: Never properly tested in its own regime. The contrarian and mathematician are right: Dresden used NMR, not DNP; the authors note DNP polarises at least ten times more nuclei and make no claim of refutation. The NMR variant is tested and not supported. Evidence also on record: Alzofon's own 1994 report (about 1.1 mg on 0.5 g, with a transient structured on a 5-10 ms scale on a 0.01 mg balance, which the contrarian flags as an instrument question) [read: Tajmar paper via deep read]; Prutchi 2019 arguing Alzofon's data show no effect (not read; [web] via Investigator); Sokol's claimed "positive correlation" between DNP and weight (his conference; poster with D. Chester; no data seen) [web: altpropulsion.com]. The calculation above (about 100% per polarised nucleus) says no theory in the folder supports a 0.22% effect. For the claim, stated fairly. The TM is an official NASA document with named method and a stated sign reversal; the fade with magnet history is what a magnet artifact predicts, but the TM says the measured field did not change, which is partly in the effect's favour (contrarian). A sign flip with geometry flip is also what a real spin force would show, so the test should be repeated with controls, not assumed an artifact. What would settle it. (i) Bismuth rotor versus a matched-mass spin-zero (e.g. non-magnetic silicon or lead) rotor in a closed housing, no disassembly between conditions, magnet-only runs with no rotor, shielded and thermally controlled. (ii) The V5 data and device from Quantum Machines. (iii) A DNP weight run at Alzofon's conditions with a no-specimen null control and temperature logging. (iv) Anyone redoing the bismuth force calculation with one consistent rule. (v) A rigid-body test: arm force versus spin rate at fixed orbital rate with a high-force load cell, against M omega^2 r. Final verdict (split). - POAMS spin-coupled force as a theory: Internally inconsistent in application (three results for one nucleus; sign convention; r = r_o at the surface), though its algebra is correct and adds nothing to its own postulate. - The claim that the 0.3 g bismuth result is "consistent with" the theory: Refuted by calculation (about 2000 times too large for the TM's own force law with thermal alignment; stated conditions: 0.2 T, 300 K). - The reported 0.3 g bismuth-rotor effect itself: Never properly tested (one uncontrolled preliminary set of runs, ordinary candidates not excluded, data withheld; not dismissed, never examined). - Gyroscope orbit prediction (Table 1): Never properly tested (contradicts rigid-body mechanics; no measurement in the predicted regime). - Alzofon DNP weight reduction: Never properly tested in its own regime; the NMR variant is Tested, not supported. - "This is the next thing to test" (her 2020 plan): Plausible, untested* as a programme; no result is in any document.

Claim 4. Mainstream null results on Podkletnov-type claims are not decisive

Agreement. As a point of method it is sound and documented: the nulls are static (Li 1997), slow and without RF (Hathaway 2003 at 400-800 rpm against 5,000 rpm), and the burden of a specification falls on the claimant (physicist). The monolayer/bilayer premise is unverified: Gallerati notes Podkletnov's 1992 disc was single-layer and the later UAH-MSFC disc bilayer, so even the original literature is unsettled [web: https://arxiv.org/pdf/2206.07574]. "Tajmar is funded by the German Air Force to publish no results" has no source in any document or in anyone's searches; the speaker is uncertain; and the Dresden group publishes detailed null and artifact analyses. Tajmar offered to test her samples and she did not send them, in her own words (distrust given as the reason) [read: interview 01:27:47]; this is evidence against a suppression reading and stays stated. Disagreement. Minor: the physicist and contrarian label the logical point "Plausible, untested"/"Plausible", the mathematician the same. What would settle it. Which protocol and layer structure Tajmar actually used (a reading of his papers), and a matched replication. The funding allegation needs a source or withdrawal. Final verdict (split). - "Nulls from modified setups do not exclude the original" (logic): Plausible, untested (the bilayer hypothesis itself is untested). - "Tajmar's nulls were irrelevant because he used a monolayer": Never properly tested (premise unverified; his Podkletnov-related work was gravitomagnetic and Poher-type, not a rotating-disc replication [web: Gallerati]). - "Funded to publish no results": Not testable as stated (an inference, attribution to her unconfirmed, no source).

Claim 5. An independent, privately funded institute is the remedy; the ether; "a mathematical equation" for six forces

Agreement. No equation, data or protocol appears in any document; the six-force statement is a conversational remark whose speaker and count (eight forces named, six "on lockdown") are uncertain. The classical ether is excluded by Michelson-Morley and later Lorentz-invariance tests [memory: high]; her statement is brief and unelaborated; a modern vacuum-structure sense is too thin to assess. The RLC/speed-of-light remark is true in form (resonance omega = 1/sqrt(LC); c = 1/sqrt(mu0 epsilon0)) and a coincidence of structure, not evidence for an ether [memory: high; read: lecture 01:04:08]. Institute facts: Delaware public benefit corporation (website); HoloChron LLC (000524283, formed 2018-07-16, dissolved 2022-12-28) and HoloChron Holdings LLC (000524282, formed 2018-07-16, dissolved 2020-12-27) in Alabama [web: https://al.ltddir.com/companies/holochron-llc/, a mirror of the state register]; a SAM entity dated 2018-12-17 per the Sentinel (not verified at SAM.gov). No Institute laboratory, grant, data or patent is documented; no patent by her (the one Eskridge patent, US7808353B1, is the plasmoid-propulsion patent, assigned from NASA to Richard Eskridge 2018-10-03) [web: Google Patents]. Which is stronger. There is nothing to weigh on the physics; the funding thesis is a business position outside the verdict scale. What would settle it. The Institute's Delaware and SAM filings, HoloChron annual reports, and the Sept 2020 C&D text. For physics, the equation itself, if it exists. Final verdict (split). - "Six forces, anti-gravity the sixth, with an equation": Not testable as stated (no equation shown; attribution to Amy uncertain). - Classical ether: Historically superseded but instructive; a modern vacuum-structure sense: Not testable as stated. - Independent funding as the remedy: outside the scale. The record shows no public results from the Institute.

Claim 6. She was surveilled, burgled and threatened because of this work

Agreement. Not physics. The record is her own detailed 2020 account (cut charger, headphones unplugged at 36% battery, an open window, a tailing Lexus with a swapped plate, a "Mana World Holdings Trust" offer, Bluetooth attacks, "handlers", 5,000 foreign agents, the NSA list of five) [read: interview 00:35:22-00:43:54, 02:11:41-02:58:33], Sokol's second-hand "havana attacked" account [web: Anomaly Archives, interested], Milburn's directed-energy claim (his own, UAP field) [web: same], the family's statement ("marvelously intelligent", chronic pain) and her father's "Scientists die also" [web: NewsNation], and a death on 11 June 2022 ruled a suicide [web]. A text she sent a friend saying she would not have killed herself resurfaced in 2026 coverage [web: search summary]. She herself calls the NSA-list source possibly a plant and says she filled in the blanks; 5,000 agents in a city of barely over 100,000 is hearsay [read: deep read of interview]. None of the records she says exist (packet captures, logs, the Uber record, the consultant's remarks, the plate, the trust offer) has been examined by anyone; no police report is mentioned. Disagreement. Physicist and mathematician: Not testable as stated; contrarian: Never properly tested for the events, Not testable as stated for the list and agent counts. I side with the contrarian's split, because the events have concrete, checkable records; the single-source status is a reason to say where each account comes from, not a reason to discount it. What would settle it. The packet captures and logs, the Uber trip record, the plate photograph, any police report (Wikipedia says Birmingham police; other sources say Huntsville/Madison County; unchecked), the consultant's engagement, the Madison County sheriff's report, the ex-boyfriend named as co-witness. Final verdict (split). The events: Never properly tested (no records examined, nothing refuted either). The NSA list and agent counts: Not testable as stated. Nothing in mainstream physics depends on it.

Absence of evidence: where the record goes dark

A. The Institute and Amy's public work after 2020. Facts [web unless tagged]: HoloChron Holdings dissolved Dec 2020 and HoloChron LLC Dec 2022; the Institute site has no sub-pages in the Wayback CDX and goes quiet after June 2022; no paper, patent, data or test report of the Institute exists; nine TV appearances 2021-24 as president (content unseen); a reported Sept 2020 cease-and-desist citing NASA, apparently sent on her behalf when Falcon Space tried to present the DNP research (secondary social-media and Anomaly Archives accounts; text not seen). Competing explanations: 1. Legal or IP restriction (the C&D; Space Act proprietary periods). For: the TM itself records a proprietary period and that the V5 data are held by the funder [read: TM p.35]; the C&D is reported. Against: text unseen; this explains the lack of disclosure, not the lack of any tests. 2. Funding failure of a venture seeking investors. For: she said "money is always the problem" [read: lecture]; no facility or grant is documented; the LLCs dissolved. Against: no investor or filing records found. 3. Health. For: family statement of chronic pain; her 2017 pain-advocacy blog [web: Anomaly Archives]. Against: none, though it does not explain the corporate history. 4. Nothing to publish (null or ambiguous results). For: the TM's own fading effect; Dresden's null for the NMR variant; the 2015 test's core null. Against: no test data of the Institute exists to show it either way. 5. Pressure or suppression as she described. For: her own detailed account; Sokol's account; a cluster of 2026 media stories. Against: no document or physical record; the father's and family's account; the C&D, if real, was sent by her side. These are not exclusive. The records that would decide: the C&D text, the Delaware filings and Institute bank and investor records, the 2020 Falcon Space test logs, the MSFC Space Act files, the TV episodes.

B. The 2015 MSFC test. Unpublished; no report found (the NTRS check was inconclusive, a probable malformed query). Explanations: an informal exploratory run not worth reporting; a Space Act proprietary restriction (documented for the bismuth work, not for this one); redirection or classification (her reading; no evidence); records exist in an MSFC file not yet requested. Decide with a FOIA request to MSFC, the film, and Richard Eskridge.

C. Ning Li after 2002. No publications after 2002; AC Gravity licence renewed through 2018; injured 2014; died 2021 [web: Huntsville Business Journal; Sentinel]. Explanations: technical stall (an audience voice says discs cracked [read: lecture 00:46:02]); loss of funding; illness and injury; classified or absorbed into defence work (no evidence); an effect that did not hold. The record supports the third best, and leaves the others open. Decide with AC Gravity's filings, the 2001 contract record, UAH and NASA collaboration records.

D. The V5 bismuth device and data. Held by Quantum Machines (funder; stake in the result) after work stopped for lack of funding [read: TM p.35]. Decide by obtaining them.

What would settle it (one deciding step per disagreement)

1. Podkletnov: a faithful public-data replication with calibrated gravimeter (Robertson's 2025 effort, per APEC, is the candidate).
2. 2015 MSFC test: the report, traces and film; a control discharge with a warm or non-superconducting disc.
3. Bismuth: a controlled spin-zero rotor, closed-housing, no-disassembly repeat; the V5 data; one consistent bismuth force calculation.
4. Alzofon: DNP weight run at 0.33 T and 9.5 GHz with a no-specimen null and temperature logging.
5. Gyroscope: arm force versus spin at fixed orbit against M omega^2 r.
6. Tajmar protocol: read his papers for layer structure; the German Air Force claim needs a source.
7. Surveillance and the Institute: the named logs, Uber record, C&D text and filings.

Open questions

• Who spoke the second Tajmar-funding line at 01:27:47 (the first, 01:05:49, is attributed to Amy; the six-forces passage is most probably Amy; the audio would settle both)?
• Who is the first-hand voice in the lecture Q&A, and is the "France" Space Act partner Poher?
• [Withdrawn by A1: the Table 1 branch difference is a sign-convention error of mine, not an open question.]
• Did the 2024 Dresden group or anyone run DNP at Alzofon's conditions?
• Is there any bismuth-versus-control run in the V5 data?
• Is the 1973 Mansfield revision relevant to the lecture's date?
• What does the Sept 2020 cease-and-desist actually say, and who sent it?
• What do the nine 2021-24 TV appearances say about her later physics?
• Are the packet captures and logs she mentions still held by anyone?

---

QE revision (2026-10-07)

Inputs: qe.md (verdict MINOR), and the new deep reads appended to deep/ on the NASA TM (pp.25-36, alignment mechanisms and sign convention), the HAL5 slides (pp.10-12 addendum), the lecture (section 8: YBCO, Becker, the first-hand MSFC voice) and the interview (section B: speakers, Austin investor, Maxar, Chantelle, Tajmar). Tags as above. Web checks added in this revision: 2 (Google Patents US1655144; search on Wu, Ashburn and PRL 58, 908), beyond the 7 used earlier; the extra two were requested by the QE remediation list. Result: no final verdict changes scale name. Four rationale statements change, one factual claim of mine is withdrawn, and the table below records every claim.

A1. Withdrawal: TM Table 1 is reproduced by eq. 28a (my error)

My "recomputation item 4" said the table was not reproduced and the mathematician's correction was "not confirmed". That was a sign error. Appendix B defines Fmeas = Fgyro - DF [read: page images pp.50-51, per the TM deep read section 6], so the spin-coupled force is Df = F(zero spin) - F(table), positive for co-spin. I recomputed with the table's six rows (F0 = 116.32 N; co-spin 115.78 ... 98.52 N; anti-spin 116.86 ... 138.66 N) and F_N = 89.85 N (disc only) [calc, 2026-10-07]: (F_N + Df_co)^2 + (F_N + Df_anti)^2 over F_N^2 = 2.0001, 2.0001, 1.9999, 2.0000, 2.0001, 2.0000 for 10 to 60 Hz. With my earlier (wrong) sign it gives 2.0001 ... 2.2021, which is exactly what I reported. The mathematician and the QE were right; the deep read (first pass), the physicist and I were wrong. The "different spin terms for the two branches (32.1 N vs 50.2 N)" came from the same sign error; the QE reports 39.12 N for both at 60 Hz, as the Appendix B sheet prints (I did not recompute that figure). What stays: Table 1 is a correct output of the printed formula, but the formula rests on the postulates; the table mixes the disc mass (0.112 kg) and the gyro mass (0.145 kg) in its normalisation (stated openly in the TM); Euler's first law (centre-of-mass theorem) still says the arm force should not depend on spin unless a new force exists; and no experiment reached the regime (V1 spin-down, 50 N load cell). Effect on verdicts: none. Effect on the claim-3 internal-consistency case: one item is removed (Table 1 not reproducible); the remaining items (three results for one nucleus, sign convention flip between the gravity and gyroscope cases, r = r_o at the Earth's surface, Appendix A against the main text, Appendix C normalisation) stand. The physicist's Correction 6 and its Table 1 bullet should be withdrawn by the orchestrator.

A2. The other alignment mechanisms (from the TM deep read, sections 3 and 5)

The TM invokes four mechanisms, none with a number: Barnett self-magnetisation of electrons, Overhauser transfer to nuclei, gradient-induced "flipping" of nuclei, and the Wallace effect (source: US patent 3,626,605, "no independent measurement" known to the deep reader [memory: medium]). It gives no rotor speed, gradient, enhancement factor or predicted alignment [read: pp.30-33]. It also states: "even for very high magnetic fields and cold temperatures the alignment fraction is limited to a very small fraction of the total" [read: p.30] and calls 0.2 T "a very weak method of excitation" [read: p.33]. My assessment [calc and memory: high unless marked]: - Thermal polarisation of Bi-209 at 0.2 T and 295-300 K: 4.1e-7 to 4.2e-7 (deep read 4.2e-7; my Brillouin sum 4.08e-7). The reported 0.3 g on 211 g (0.142%) needs a co-spin aligned fraction of 8.4e-4 on the TM's own factor 1.69, about 2000 times thermal (about 1000 times on the anti-spin factor 3.29). Agrees with item 1 above. - Barnett/rotation: effective field omega/gamma is 2.4e-5 T at an assumed 10,000 rpm (the TM states no rpm), about 1e-4 of the applied field. Negligible. Einstein-de Haas and Barnett concern electron spin in iron; bismuth is diamagnetic. - Overhauser transfer: the enhancement requires saturating the conduction-electron spin resonance with microwaves (about 5.6 GHz at 0.2 T). The ceiling, gamma_e/gamma_n about 4,000, would give P up to about 1.7e-3, which is enough in principle to match 0.14%. The TM describes no microwave source, so without saturation the electron-nucleus coupling only relaxes the nuclei toward thermal value. Bismuth's very low carrier density (about 1e-5 per atom [memory: medium; not web-checked]) works against the scheme. - Gradient flipping: a gradient exerts a force on a moment (F = grad(mu.B)); the torque makes it precess, not align. The equilibrium value is still set by local B/T. No flip rate is calculated. - Wallace effect: no magnitude given. Conclusion: the verdict stands as conditioned. "The 0.3 g result is consistent with the theory": Refuted by calculation for the means the TM describes (static 0.2 T, room temperature, rotation, no microwaves); the TM supplies no number that lifts the alignment above thermal. The condition is now stated in both directions: a saturated Overhauser process with microwaves could in principle reach the needed size, and the TM does not say one was present. That is a reason to run the test described below, not a reason to credit the TM's mechanism. The claim is not refuted for every conceivable spin-weight coupling.

A3. Material kept in the record for claims 2 and 3 (retention)

Claim 2, the first-hand voice (lecture section 8.4-8.5) states both sides, in order: glass dewars blew up and the work moved to Walmart buckets (00:30:10); "we got an excess of 500 Gs, but we never reproduced the waves" (00:49:38); a second superconductor under the dewar floated a magnet and "when we fired the Boyer device the little magnet went far away" (00:50:09-00:51:12); "we told management about this" (00:51:12); "when you're doing a pulse power experiment with 20,000 amps at 2,000 volts, getting a signal is never a problem ... it's not a question of getting the signal, it's a question of getting the right signal"; "with a double Faraday cage, we did not reproduce Boyer's signals"; "However, we did see a whole lot of really strange effects. Like the tossing of the magnet and forces coming off the disc" (00:51:42); "if it was a thermal explosion between the layers of the disc, it would have destroyed the disc, but you pull it out and it was pristine"; a 1,000,000 frames-per-second film of "a wave coming off" an emitter; "I was impressed with it" (00:52:13-00:52:44). The same voice recalls the 1990s Podkletnov-era excitement at Marshall ("this could be a revolutionary thing", 00:45:01) and says "we did make some smaller ones" of Ning Li-type discs (00:46:32). The pristine disc is the account's own argument against a thermal-explosion explanation and is not rebutted in the text. The mundane candidates I listed (flash boiling, a transient field of about 0.08 T at 5 cm from 20 kA [calc], steam shock, pickup) are not excluded by the account either: no control discharge with a warm or non-superconducting disc is reported, and no magnet mass, distance or timing. The voice is not identified; the best-supported reading is that it is Richard Eskridge, from adjacency to "my dad did that" (an inference; the audio is needed). Verdict unchanged: Never properly tested.

Claim 3, the TM's own arguments (TM deep read section 4, [read: p.33]): (i) the motor is a three-phase pulsed-DC type and "does not have a reversal of DC current which might explain a reversal of forces"; (ii) "The external magnetic field of the magnets was measured from test to test and did not change"; (iii) the effect "returned with equal force" with a new device and new magnets and was "later suspected" to be a rare-earth magnet phenomenon. Limits stated by the deep read: argument (i) covers the motor supply currents only; in V3 the co-spin run followed disassembly and flipping of the rotor and magnets, so magnet orientation relative to the scale, the Earth's field and the motor also changed between runs; the instrument and point for the field measurement are not named; no run count, no control rotor (spin-zero material), magnets-only, stalled-rotor, perpendicular-axis, temperature or airflow run is reported; the only comparison is anti-spin against co-spin. The TM's own words: first loss "did not inspire much excitement because this weight loss could be due to thermal buoyancy or aerodynamic forces"; data "rudimentary and preliminary". Facts of the V4: the drawing of the rotor cup is dated 2016-12-14, a month after the V4 photograph (2016-11-14), unexplained. Verdict unchanged: Never properly tested for the 0.3 g effect.

A4. Absence of evidence A and "What would settle it": new leads (interview deep read B6, B7, B9)

• Maxar (interview 00:57:56). She reads her calendar: "negotiating with maxar about selling them a plasma thruster patent on the 17th of december" [2019]. Whose patent, number and outcome are not stated. The Investigator's find, US7808353B1 (plasmoid-thruster coil patent, assigned from NASA to Richard H. Eskridge on 2018-10-03 [web: https://patents.google.com/patent/US7808353B1/en]) is a candidate; the link is my inference, not documented. It adds a documented commercial activity of the family to the "funding" and "legal/IP" explanations in A: the Institute was trying to sell an asset, and a patent assignment out of NASA in 2018 is the kind of record that would explain a NASA-related cease-and-desist. Records that decide: the assignment file and any Maxar correspondence.
• Austin investor and "Hal" (00:47:14). She says her father "got poached" out of NASA by "this weird ass investor ... in austin texas ... a real estate investor", who brought Hal [Puthoff] to the family house three times; she sat with him once and he talked about "ultra terrestrials". The investor is unnamed and no year is given. The TM says Nelson's Space Act partners included Austin, Texas [read: TM p.37], and Quantum Machines (CEO Chris Milam) funded the bismuth work; the TM gives no location for Quantum Machines. Whether the investor is Quantum Machines is unknown; I did not check. Her later opinion of Puthoff ("extremely evil", no incident cited, 03:01:06) is hers.
• Father's long interest in Podkletnov. "In 2004 my dad tried to get me to do the podkletnov experiment for the science fair" (01:31:33; 01:26:06 says "junior in high school"): a first-hand statement, no data, that the 2015 MSFC test followed a decade of family interest.
• Prototypes "melted down for scrap" (01:16:15). A general claim of cancelled NASA prototypes sold at divestment auction, no device, date or name given though she offers to supply names. Records that would test it: MSFC property disposal and GSA auction records. Uncorroborated; stated as hers.
• Becker and the DIA loan (lecture 00:21:27-00:22:32). Her account: the UAH thesis by Robert Becker on quantised gravity has been checked out twice, in 1999 under a "DIA inter departmental library loan" (caption wording) and by her. How she saw the loan record is not stated. The 2018 website lists Becker as the Institute's technical adviser [read, per the QE]. A UAH library circulation record would settle it.
• YBCO story (lecture 00:19:51, 00:28:36). Her account: YBCO was invented at UAH by a grad student and a patent was "stolen by another university" the next day. "Jim Ashworth" is the caption's name; [web: https://www.aps.org/apsnews/200307/prl-2.cfm and https://ethw.org/First-Hand:Discovery_of_Superconductivity_at_93_K_in_YBCO:_The_View_from_Ground_Zero] show Wu and his graduate students Jim Ashburn and T. J. Torng at UAH, with Chu, in PRL 58, 908 (1987). So the name is almost certainly Ashburn (confirmed by search summary; the identification of the caption name remains my inference). The "stolen patent" has not been checked and is not a physics claim; the priority dispute between Houston and UAH is in the YBCO literature [memory: low, unchecked].
• Chantelle is, by the interview, "chief research officer of geometric energy in calgary alberta" and her business partner; the 2018 website's research director is probably the same person (QE; surname not in the interview). She was present at the 2018 Estes Park confrontation with Tajmar and is a possible second witness.

A5. Attribution corrections (claim 4 and claim 5) and tag fixes

• Monolayer/bilayer argument: Amy's own, first person ("chantelle and i basically sprung on him ... right here in the protocol it says you used a mono layer not a bilayer", 01:25:34-01:27:47; speaker high confidence per the interview deep read). My "speaker not certain" note was too cautious.
• German Air Force funding: Amy states it herself at 01:05:49 ("he's funded by the german air force all he ever publishes is no results"), with no grant, document or Tajmar statement behind it; the repeat at 01:27:47 is of uncertain speaker. It is her assertion, uncited. It is weighed against her own report that Tajmar offered to test her samples and she refused (the same passage).
• "Six forces": attributed to Amy, medium-high confidence: "my group has a mathematical equation ... we have six forces on lockdown" (00:20:20); Mark Sokol says he does not understand it (00:29:38). The first four forces are not captured, the eighth is never named, and no equation is shown. My earlier note that the speaker "may be Mark Sokol or Jeremy Rys" is withdrawn as unlikely. Mark's own account of the airport conversation with her father is where the dynamic nuclear orientation neutron-beam design comes from (hydrogen, not bismuth).
• Tag fix: "I don't know what happened" about Ning Li is [read: lecture 00:18:16], not slide p.25.
• Tesla slide (new item, claim 1): slide p.12 says "Patent 1,655,144: a 'space drive' or anti-electromagnetic propulsion system". US1655144 is a grain-treating machine (Hallgarth and Chandler, 1928) [web: https://patents.google.com/patent/US1655144A/en]; the number is wrong or does not belong to Tesla. The source on the slide is William Lyne's "Occult Ether Physics" (an interested popular book); the saucer drawing on slide p.11 is unattributed. The deep reader's [memory: low] recollection that Tesla's 1928 flying-machine patents are 1,655,113 and 1,655,114 was not checked by me. This is an error on her slide; it does not change a verdict (no verdict depends on a Tesla device).
• Slide page: "Torr's work never made public" is slide p.28, not p.27 (facts.json fix).

Verdict table (changes marked)

Corrected claim list (carry forward)

1. Claim 3, Table 1: reproduced by eq. 28a with Fmeas = Fgyro - DF (F_N = 89.85 N; spin term 39.12 N at 60 Hz per the sheet). The arbiter's first statement that it was not reproduced is withdrawn.
2. Claim 3: the 0.3 g result needs about 2000 times thermal alignment; the TM names four alignment mechanisms with no number for any; Overhauser transfer would need microwave saturation that the TM does not describe.
3. Claim 3: add the TM's motor-phasing argument, the unchanged-field measurement and the V3 reassembly confound to claims.md.
4. Claim 4: the monolayer/bilayer argument is Amy's; the German Air Force line is her assertion with no source (01:05:49), the repeat at 01:27:47 of uncertain speaker.
5. Claim 5: the "six forces" remark is most probably Amy's; no equation shown.
6. Claim 2: add the pristine-disc argument, "strange effects" summary, "we told management" and the 1,000,000 fps film as the account's own for-side; the voice is unidentified (probably, not confirmed, her father).
7. Claim 1: Tesla slide patent number 1,655,144 is a grain-treating machine; Podkletnov-era "Jim Ashworth" is probably Ashburn of UAH.
8. Absence of evidence A: add the Maxar negotiation, the NASA-to-Richard-Eskridge patent assignment (link inferred), the Austin investor, the Becker loan record and the prototype-scrap claim as records to obtain.
9. Tag fixes: "I don't know what happened" is lecture 00:18:16; Torr "never made public" is slide p.28.
10. Not changed by this round and still unchecked: spin-gravity torsion-balance bounds [memory: medium]; Tajmar's actual protocol and layer structure; the 1973 Mansfield revision; the "resurfaced" 2026 text (seen only in search summaries); the E3 illustration in the primer (A6: it combines eqs 28b and 31 orientation by orientation; I did not re-derive it here, so the primer should either state it as an illustration of the "Internally inconsistent" verdict with its assumption or drop it).