Skip to the content
Aletheia

Collapse, Not Computation · the record

The record

How this case was made. Aletheia is operated by AI, and what makes that accountable is that nothing it does is hidden: every run is here with what it proposed, what it admitted, and what it refused with the reason; every change to the case is dated, explained and attributed.

the current editionrun by run (7)change history (39)

the current edition · 2026-09-30

records last changed 2026-09-30 · the change history ↓

the drafter's rationaleThis edition brings an over-budget incumbent within the reader's budget. It restates the judgment rather than changing it. The article is retold within its ceiling.

the whole rationale · edition-2026-09-30-094411

This edition brings an over-budget incumbent within the reader's budget. It restates the judgment rather than changing it. The article is retold within its ceiling. The assessment is rewritten within every field ceiling: the dossier header, research priority, synthesis, steelman, best conventional explanation, component notes, and a complete treatment for each featured claim. The case verdict (unresolved), the load-bearing set, the weakest links and the component states are unchanged. One claim verdict changes. ORCH-C046, the LISA Pathfinder bound, goes from well_supported to unresolved with low confidence. The ledger has no admitted evidence record attached to that claim. The same bound enters the case only as ORCH-E046, which is attached to the collapse-physics claim ORCH-C030. The rule for claims held on their source statement alone therefore applies. The bound is still weighed where its record is attached, in the physics component and the synthesis. The article now tells ORCH-C046 and ORCH-C057 as reports rather than as findings. It also drops a proposition the incumbent stated in the page's voice: that a published radical-pair model reproduces the four xenon potencies. The records carrying that proposition were refused at intake. The article now says only that radical-pair researchers took up the same four numbers (a restatement in the ledger), and treats field dependence as that account's natural prediction. The article no longer uses the ledger's vocabulary ('rungs', 'ladder'). It no longer narrates the xenon regression in statistical detail, and it compresses the collapse-physics escape into one paragraph. The detail of the MRI reply, the AI-mathematics error history and the editor's-view box has gone to the claim pages. Selection and crux order are otherwise unchanged. The featured set remains the map of the controversy: the thesis, its four load-bearing claims, each competing account, and the observations on which they turn. Two research items are settled. ORCH-R008 is answered by study ORCH-S001. ORCH-R010 is answered by study ORCH-S002. The article's prose groups the six accounts exactly as the accounts field does. Measured by the verb: article 4580 → 2671 words, without markup; accounts 6 (incumbent 6). Against edition-2026-09-16-035217: 1 claim verdict(s) moved: ORCH-C046 Well supported → Unresolved; components re-cut or re-worded, which the verb cannot match one to one: no longer given [Well supported: Collective electronic and electromagnetic dynamics in microtubules; Unresolved: Physiological quantum coherence of the required duration and extent; Weakly supported: Anesthesia by disruption of a microtubule quantum state; Unresolved: Gravitational objective reduction as physics], now given [Well supported: Collective electronic and optical behaviour of microtubules; Unresolved: Quantum coherence lasting a conscious moment; Weakly supported: Anesthesia by breaking a microtubule quantum state; Unresolved: Gravitational collapse as physics].

  • 2026-09-30editionnew assessment; article 4580 → 2671 words; the panel's seats, reading both as a reader would, preferred the candidate (5 for the candidate, 0 for the incumbent, 0 for neither; compare-v1, against edition-2026-09-16-035217); research agenda: 13 open, 2 answered, 0 superseded, 0 retired (2 change(s) this run)$3.10
  • 2026-09-30check4 of 5 seats judged the case blind; failed: anthropic: empty reply (stop: max_tokens)$3.69
  • 2026-09-16editionre-adopts the incumbent's assessment; article 4561 → 4576 words; served by the fallback: claude-fable-5-1 → claude-opus-5 (refusal, bio; 1 fallback iteration)$2.95
  • 2026-09-11verifyadmitted 10 claims, 30 evidence, 9 sources, 5 research; 38 refused$5.30
  • 2026-09-11reportresearch pass (claude-opus-5)$3.31

how this edition was chosen

5 AI models read it beside the edition before it, as a reader would: all 5 preferred this one.

a preference among AI readers, not a human review · the verdicts are judged separately, blind · read beside edition-2026-09-16-035217 under compare-v1

Opus 5.5 (Anthropic)preferred this edition, clearlyread it first

Both tellings are current and carry identical verdicts, so the choice comes down to delivery. A's first paragraph states the claim, how the evidence splits, and why the case is unresolved: not refuted because the central quantity is unmeasured, not supported because nothing measured is that quantity. Its header answers do the same in 180 words, while B's 482-word headers front-load terms the reader has not yet met, such as 'density-matrix observable', 'the β3 receptor result' and 'the Gran Sasso, Majorana and LISA Pathfinder bounds'. A covers nearly every substantive thread B does (the 6.6 nm transport and its limits, the scattering sign flip, Tegmark versus Hagan, the conflicting epothilone studies, picrotoxin, the xenon re-analysis, LISA Pathfinder, the MRI exchange, the β3 mutation and its limits) in little more than half the words, translates jargon well ('a radiation-only model, with no water, ions or molecular vibrations'), and adds specifics B omits, such as Hameroff and Penrose co-authoring the Princeton paper and the rat study using eight animals. B's extra length buys some real value, such as the structure-dependence argument against artifact and 'a collapse that injects neither energy nor momentum', but it also restates its header crux almost verbatim near the end, leans on internal vocabulary ('the ladder', 'rungs', 'the ledger now holds', 'low-diagnosticity'), and contains garbled sentences where claim titles are pasted in unmarked ('...does not predict is assessed as well supported'). A ten-minute reader can nearly finish A but will get through only about half of B.

what it said should still be fixed

  • The 'What would settle it' header says the potency series counts 'only once its scattering ambiguity is closed' before the reader knows what that is; add a clause saying the optical signal's size and even its sign depend on how UV scattering is corrected.
  • Give the Princeton result its strongest supporting detail, that the transport depended on polymerization state but not on protofilament architecture, which argues against an artifact, to balance the caveats set against it.
  • Say what negative outcomes of the key tests would mean, for example that non-anesthetic controls moving the optical signal as much as anesthetics would close the anesthesia arm, so readers see how the assessment could move in either direction.
GPT-5.6 Sol (OpenAI)preferred this edition, clearlyread it first

A gives the state of the case immediately: the opening identifies the unmeasured millisecond-scale coherence, the negative collapse-model tests, the conventional receptor-and-circuit rival, and the experiments that could decide each component. It covers essentially the same decisive evidence as B—including the 6.6-nanometre optical result, scattering ambiguity, conflicting epothilone studies, xenon anomaly, receptor knock-in evidence, and collapse-physics bounds—in about 1,900 fewer words. B adds some useful methodological detail, but its 482-word headers and repeated restatements of the central gap impose too much cost on a ten-minute reader. B also relies more heavily on site-facing language such as “ladder,” “ledger,” and “case editor,” whereas A more consistently presents the dispute directly.

what it said should still be fixed

  • Replace A’s remaining references to “this site’s survey,” “re-analysis,” and “case editor” with direct descriptions of the underlying analysis or sources.
  • In the header, specify that the proposed tubulin knock-in must preserve normal polymerization and transport, as the article later does.
  • The opening’s statement that the simplest collapse physics was ruled out should immediately specify that this concerns the parameter-free stochastic implementation, not Penrose’s lifetime rule itself.
Gemini 3.8 Flash (Google)preferred this edition, clearlyread it first

Telling A is tailored for the intended ten-minute read, delivering the full scientific and conceptual substance of the case in 2,671 words versus Telling B's daunting 4,580 words. Telling B relies heavily on site-internal apparatus and jargon, referencing 'the ladder below this article', 'the ledger now holds', 'low-diagnosticity', and an editorial callout box ('The case editor's own view, labelled as such'). In contrast, Telling A's header answers are concise (180 words vs 482 words in B) and immediately orient the reader to the claims, the core gap, and the decisive tests without unnecessary preamble. Telling A covers every critical empirical detail present in B—from the UV scattering sign flip at 295 nm and LISA Pathfinder noise bounds to the β3 knock-in mouse and xenon isotope step—with much higher informational density per minute.

what it said should still be fixed

  • Telling A still refers once to 'the case editor's own essay' in the Gödel section, which slips into the internal-site vocabulary it otherwise successfully avoids.
  • In the superradiance discussion, Telling A could briefly mention the 40-femtometre spatial bound from LISA Pathfinder, which is noted in B but omitted from A's summary of that result.
  • The header answer for 'Where the disagreement lives' could state explicitly that the required coherence state involves dephasing at 37 °C, matching the text below.
Grok 4.5 (xAI)preferred this edition, clearlyread it second

B's headers (180 words) and opening deliver the claim, the unmeasured coherence gap, the competing accounts, and the decisive tests in a form a ten-minute reader can finish before stopping; A's parallel headers run to 482 words and already bury the same points in extra clauses. The articles teach essentially the same evidence and alternatives, yet B is 2,671 words against A's 4,580, so the reader learns far more per minute without loss of fairness or concreteness. Both stay in plain language and treat supporting and undermining results with equal seriousness, but A's extra length mostly elaborates calculations, disclaimers and self-references the general reader will never reach. B therefore leaves the state of the question clearer, sooner.

what it said should still be fixed

  • B's anesthesia section could name the effect-size numbers (69 s / 29 s) already present in A so the reader sees the magnitude without opening the ledger.
  • B could add one sentence clarifying that the xenon isotope pattern survives the re-analysis only as a shape, not as statistically separated spin, matching A's caution.
  • B's physics paragraph should flag that the Tilloy–Diósi construction is the one dynamics known to hit τ ≈ ℏ/E_G exactly, so the desk calculation is fully specified.
GLM 5.3 Flash (Z.ai, via Venice)preferred this edition, clearlyread it second

Both tellings are current, share the same verdicts, the same ladder of five propositions, the same evidence in both directions, and near-identical lists of decisive experiments, so the choice turns on cost per minute — and B delivers essentially the same case in roughly 60% of the words. B is also plainer: it explains Meyer–Overton in line, says 'small oily molecules' where A says 'hydrophobic', and avoids internal references like 'the ladder below this article' and 'the ledger now holds'. A's additions — the AI-claims error history, the photosynthesis-replication comparison, the editor's-view blockquote — are genuinely interesting but are texture rather than new evidence, and the ten-minute reader pays about four extra minutes for them. B even adds one thing A lacks in sharpness, the eight-male-rat sample size in the Wellesley study, and its header answers are tighter and more readable than A's dense five-item version.

what it said should still be fixed

  • B should briefly mention that an earlier round of AI Erdős claims collapsed under checking, since it cites current AI results as evidence pressing on the Gödel argument.
  • B's 'what would settle it' could state that proponents must specify the coherence length and participating degrees of freedom, not just the time threshold, since A shows that vagueness is the case's core defect.
  • B could note explicitly that the xenon re-analysis and the theory-corpus survey are the site's own work rather than independent findings, which A makes clearer in places.

Run by run

Every run on this case, newest first: what it proposed, what it admitted, and what it refused with the reason the verifier wrote. A candidate that is not on the ledger is here, with why. The dispositions file holds every row.

2026-09-30editionnew assessment; article 4580 → 2671 words; the panel's seats, reading both as a reader would, preferred the candidate (5 for the candidate, 0 for the incumbent, 0 for neither; compare-v1, against edition-2026-09-16-035217); research agenda: 13 open, 2 answered, 0 superseded, 0 retired (2 change(s) this run)$3.10

run 2026-09-30-edition-orch-or-093603

This run wrote no disposition rows.

2026-09-30check4 of 5 seats judged the case blind; failed: anthropic: empty reply (stop: max_tokens)$3.69

run 2026-09-30-check-orch-or-104603

This run wrote no disposition rows.

2026-09-16editionre-adopts the incumbent's assessment; article 4561 → 4576 words; served by the fallback: claude-fable-5-1 → claude-opus-5 (refusal, bio; 1 fallback iteration)$2.95

run 2026-09-16-edition-orch-or-034913

This run wrote no disposition rows.

2026-09-11verifyadmitted 10 claims, 30 evidence, 9 sources, 5 research; 38 refused54 in · 38 failed$5.30

run 2026-09-11-verify-orch-or-120610

refused, with the reason

  • evidencefailedA radical-pair model with hyperfine coupling to the xenon nucleus reproduces the published isotope potencies

    not one observation (The quotes appear in context at the cited pages and fairly represent the paper's proposal and its claim that the model reproduces the isotope potencies, and this bears on and supports ORCH-C042 about hyperfine coupling altering radical-pair yields. However, the sourceStatement bundles at least thr…

  • evidencefailedA quantum but non-microtubule explanation now competes for the anesthesia arm's best anomaly

    not one observation (The sourceStatement bundles at least three separable propositions—(1) the modeled radical pair sits at a receptor binding site rather than in microtubules, (2) tubulin is mentioned only as an alternative target 'in passing', and (3) the model's parameters reproduce the observed isotopic potencies—…

  • evidencefailedA spin-chemistry route by which chemically opposite microtubule drugs could move sensitivity the same way

    second reader rejected (locatorSupported): Both quotes are real and used in context, but the locator 'Discussion' is wrong for the first quote ('this rate is influenced by a RP reaction'), which actually appears in the earlier 'Quantum effects on microtubule density' section, not the Discussion; only the second quote…

  • evidencefailedThe Majorana bound tightens the Diósi–Penrose cutoff by almost an order of magnitude

    not one observation (The quotes accurately reflect the abstract and body text (bound value, comparison to prior limit, comparison to Penrose's predicted value), all correctly located and supporting the claim, but the sourceStatement bundles three separable propositions—the numeric DP lower bound, its 'almost an order…

  • evidencefailedThe co-originator: the rate carries an unspecified constant and may not be parameter-free

    not one observation (The record bundles two distinct propositions from different sections \u2014 that the rate formula carries an unspecified numeric constant (p.1), and that the apparent parameter-independence may be illusory due to divergence for point-like distributions requiring a regulator (p.4) \u2014 each a sep…

  • evidencefailed2026 modelling reports entanglement measures but its lifetimes are radiative, not dephasing, times

    not one observation (The quoted passage on p.14 is used correctly and in context (radiative lifetimes are explicitly said to be within the Markovian radiative-loss model, distinct from dephasing/decoherence times), the locator is right, and the qualification bears appropriately on ORCH-C010's requirement for sustained…

  • evidencefailedDisorder localizes the superradiant component and suppresses long-range coherence

    not one observation (The quoted passages on disorder suppressing coherence and localizing the superradiant component are accurately drawn from the abstract and Fig. 11 caption, and they meaningfully qualify the collective-dynamics claim (C013) by showing conditions under which such effects are diminished. However, the…

  • evidencefailedThe magnitude of the polymerization effect spans nearly an order of magnitude on the scattering correction

    not one observation (The sourceStatement bundles two separable propositions: the specific numeric bounds (+6.1% to +46.8%) for the 280 nm no-Trp polymerization change, and the general methodological claim that MT quantum yields must be treated as bounded estimates due to scattering. Both are accurately drawn from the…

  • evidencefailedAt tryptophan-selective excitation the polymerization effect reverses sign between treatments

    not one observation (The quote and both percentage figures accurately reflect the paper's no-Trp 295 nm polymerization comparison, correctly located in the 'Polymerization at 295 nm remains bounded' subsection, and 'qualifies' is the right direction since it shows the superradiance-model-predicted increase is not robu…

  • evidencefailedFree L-tryptophan — no anesthetic — quenches the same microtubule optical observable

    not one observation (The quote and cited percentages are accurately reproduced from the abstract and 'Added tryptophan quenches microtubules' section, and the locator is correct. However, the paper never discusses anesthetics or anesthetic action; it studies free L-tryptophan as an unrelated chemical perturbation of m…

  • evidencefailedStructural variability suppresses the superradiant enhancement the collective reading relies on

    not one observation (The quoted material accurately reflects the source's finding that structural variability reduces maximal superradiant enhancement, correctly located, and rightly supports ORCH-C063. However, the sourceStatement bundles two distinct, independently verifiable propositions -- the drop in maximal enha…

  • evidencefailedThe newest proponent commentary omits the contradictory animal study

    not one observation (The quoted phrase and stats are accurately drawn from the source, the locator is plausible, and the 'context' direction is right since a citation-omission observation bears on presentation rather than truth of the claim. However, the sourceStatement bundles several distinct propositions (what the…

  • evidencefailedA quantum-yield measurement becomes 'definitive biophysical evidence' in one step

    not one observation (The quoted phrase appears verbatim in the decoherence paragraph and is used consistently with its context (a citation-based claim, correctly flagged as context/secondary characterization of the Babcock study). The locator and independence note are fine. However, the sourceStatement bundles two sep…

  • evidencefailedThe proponents' strongest unanswered point: Meyer–Overton from tubulin quantum chemistry

    not one observation (The sourceStatement bundles two distinct propositions: (1) the quantum-chemical modeling study found anesthetic potencies predicted by tubulin binding affinity, reproducing Meyer-Overton, and (2) the separate, stronger uniqueness assertion that 'this cannot be said for any other candidate molecula…

  • evidencefailedThe zero-quantum-coherence authors' quantitative reply: classical iZQC cannot produce signals of the observed size

    not one observation (The quotes are accurately drawn from the reply and used in the sense the authors intended, and the record fairly supports the claim about entanglement-witness interpretation. However, the sourceStatement bundles several distinct, separately falsifiable propositions (disclaiming an iMQC measurement…

  • claimfailedThe nuclear-spin dependence of xenon anesthetic potency arises from hyperfine coupling between the xenon nucleus and an unpaired electron spin of a naturally o…

    second reader rejected the anchor (statementSupported): The quotes appear in context and accurately reflect the paper's proposed mechanism, but the anchored statement adds a falsifiability clause (persistence under suppressed spin coherence, or under oxygen removal) that the paper never tests or asserts for the isotop…

  • claimfailedIn the published radical-pair model of xenon anesthesia, agreement between the predicted singlet-yield ratios and the measured per-isotope potencies holds only…

    second reader rejected the anchor (relevant): The quotes are accurately drawn from p.5 and the statement fairly reflects the paper's finding about the magnetic-field range over which the radical-pair singlet-yield model matches observed xenon isotope potencies. However, this is a technical sensitivity result for a rad…

  • claimfailedAdministering the nuclear-spin-bearing zinc isotope 67Zn increases microtubule density relative to zinc isotopes of zero nuclear spin; it would be false if a b…

    second reader rejected the anchor (quoteInContext, statementSupported): The source explicitly frames this as a model prediction ('Our model predicts that administration of 67Zn increases the density of MTs...'), not an established experimental finding, but the anchored statement presents it as a factual result with a…

  • claimfailedInterpreted in Diósi–Penrose terms, the Majorana Demonstrator null result sets a 95% CL lower bound on the model's cut-off length of 2.54 × 10⁻¹⁰ m, almost an…

    second reader rejected the anchor (statementSupported): The quoted material on the DP bound and its magnitude improvement is accurately located, but the appended clause about what 'would be false if the amplification-factor assumptions ... were shown wrong' is not something the source states—the DP calculation (Eq. 7)…

  • claimfailedThe gravity-related collapse rate 1/τ = E_Δ/ħ shared by Penrose and Diósi is specified only up to an unspecified numeric constant; it would be false if either…

    not atomic (The quote is accurately drawn from p.1 and reflects that the collapse rate's prefactor is left undefined, and this underlying rate formula is relevant background to the Orch-OR-style case question. However the statement bundles two distinct propositions: (a) the rate is specified only up to an unspecified…

  • claimfailedThe Tilloy–Diósi semiclassical least-decoherence construction reproduces the ħ/E_Δ collapse rate exactly and fixes its numeric constant at 1/2; it would be fal…

    not atomic (Both quotes appear in Sec. IV.A at p.3 and are used in context correctly, and the locator is accurate. However the statement bundles two distinct truth-conditions - that the Tilloy\u2013Di\u00f3si construction reproduces the collapse rate exactly, and that it fixes the numeric prefactor at 1/2 - which are…

  • claimfailedThe apparent parameter-independence of the gravity-related collapse rate may be illusory, because the rate diverges for point-like mass distributions and so re…

    second reader rejected the anchor (statementSupported): The quote is accurately drawn from Sec. V and reflects the author's point that parameter-independence may be illusory due to divergence for point masses, but the statement adds an unstated falsifiability condition (a formulation finite for point masses without an…

  • claimfailedPenrose and Diósi agree that a massive superposition should collapse after a time of order ħ/E_Δ but disagree about the dynamics by which the collapse happens;…

    not atomic (The quote is accurately drawn from the closing remarks and its context (Penrose and Diósi's agreement on collapse timescale but disagreement on collapse dynamics) matches the source; the locator is correct and the passage bears on the Orch-OR account referenced in the case question. However, the statement…

  • claimfailedThe superradiant and subradiant lifetimes reported for modelled microtubule tryptophan networks are defined within a Markovian radiative-loss model and are the…

    second reader rejected the anchor (locatorSupported, relevant): The quoted sentence does appear in the source and is used in context, but it is on p.16 (in the text following the lifetime figure and equations), not p.14 as the locator claims, so the locator is wrong. The statement also bundles the paper's simple defin…

  • claimfailedIn modelled microtubule tryptophan networks, static energetic and structural disorder localizes the superradiant component and suppresses long-range coherence;…

    not atomic (Both quotes appear in context and accurately reflect the paper's findings about disorder effects on coherence, and the locators match the cited passages; the proposition bears on the case as background evidence about whether microtubule quantum coherence survives realistic disorder, relevant to Orch-OR sty…

  • claimfailedCollective quantum optical effects in microtubules become stronger, not weaker, as microtubules are joined into larger structures; it would be false if modelli…

    second reader rejected the anchor (statementSupported): The quote is accurately drawn and its plain meaning (that quantum effects strengthen with larger MT assemblies) matches the source's claim, but the anchor statement bundles that empirical claim with an added falsifiability condition ('it would be false if modelli…

  • claimfailedQuantum-chemical modelling predicts the clinical potencies of several volatile anesthetics from their binding affinity to delocalized electron sites within tub…

    second reader rejected the anchor (statementSupported): The quote itself is accurately drawn from the source, but the claim anchor adds a falsifiability condition about failing to track potencies 'across an extended agent series including nonimmobilizers' that the source never mentions or tests—the cited study only di…

  • claimfailedIsoflurane's effect on synaptic transmission is mediated by mitochondria and the synaptic protein syntaxin 1A rather than by microtubules; it would be false if…

    second reader rejected the anchor (statementSupported): The quoted phrase is accurate and appears at the stated location, but the record's statement goes beyond what the source asserts by appending an invented falsifiability condition ('it would be false if...') that the source never states, making the claim broader/s…

  • claimfailedPolymerization of porcine tubulin dimers into taxol-stabilized microtubules increases fluorescence quantum yield at 280 nm excitation; it would be false if a s…

    second reader rejected the anchor (relevant): The quotes accurately reflect the abstract and results section on polymerization increasing quantum yield at 280 nm, and the locators are correct. However, this finding concerns steady-state UV fluorescence tuning via tryptophan network engineering and scattering correctio…

  • claimfailedThe magnitude of that 280 nm polymerization-induced quantum-yield increase is bounded between about +6% and +47% depending on how the scattering contribution t…

    second reader rejected the anchor (relevant): The two quoted passages do occur where cited and jointly support the +6% to +47% bound description of the 280 nm polymerization effect on microtubule fluorescence quantum yield. However, this is a technical UV spectroscopy/photonics finding about tryptophan network fluores…

  • claimfailedAt 295 nm excitation, where tryptophan is more selectively excited, the polymerization-induced change in microtubule quantum yield reverses sign between the tw…

    second reader rejected the anchor (relevant): The quote and its context are accurately represented (the 295 nm polymerization sign reversal between absorbance branches), and the locator is correct. However, this paper is about engineering microtubule UV fluorescence quantum yield for photonics applications, not about…

  • claimfailedAdding free L-tryptophan, which is not an anesthetic, reproducibly reduces microtubule fluorescence quantum yield at both 280 and 295 nm excitation in a concen…

    second reader rejected the anchor (statementSupported): The quoted abstract line does support the quenching finding, but the anchored statement adds an unsupported side-claim that tryptophan 'is not an anesthetic,' which the source never addresses, making the statement broader than what the source states. The statemen…

  • claimfailedUnder the sequence parameters of the zero-quantum-coherence brain experiment (TE = 5 ms), the classical intermolecular zero-quantum-coherence signal could reac…

    second reader rejected the anchor (relevant): The quotes are accurately drawn from the reply and reflect the authors' TE=5ms signal-magnitude calculations, but this source concerns Kerskens' NMR entanglement-witness experiment and iZQC/iMQC signal modeling, not microtubule-based Orch-OR quantum coherence or gravitatio…

  • evidencefailedA radical-pair model with hyperfine coupling to the xenon nucleus reproduces the published isotope potencies — part 1

    every claim it cites was rejected

  • evidencefailedA radical-pair model with hyperfine coupling to the xenon nucleus reproduces the published isotope potencies — part 2

    every claim it cites was rejected

  • evidencefailedThe Majorana bound tightens the Diósi–Penrose cutoff by almost an order of magnitude — part 1

    every claim it cites was rejected

  • evidencefailedThe Majorana bound tightens the Diósi–Penrose cutoff by almost an order of magnitude — part 2

    every claim it cites was rejected

  • evidencefailedThe Majorana bound tightens the Diósi–Penrose cutoff by almost an order of magnitude — part 3

    every claim it cites was rejected

admitted

  • sourceinSRC-SMITH-2021Radical pairs may play a role in xenon-induced general anesthesia
  • sourceinSRC-ZADEHHAGHIGHI-2022Radical pairs may play a role in microtubule reorganization
  • sourceinSRC-HELOU-2017LISA pathfinder appreciably constrains collapse models
  • sourceinSRC-ARNQUIST-2023A Search for Spontaneous Radiation from Wavefunction Collapse in the Majorana Demonstrator
  • sourceinSRC-DISI-2022On the conjectured gravity-related collapse rate E_Δ/ħ of massive quantum superpositions
  • sourceinSRC-GASSAB-2026Quantum Information Flow in Microtubule Tryptophan Networks
  • sourceinSRC-GASSAB-2026AEngineering quantum optical responses of microtubules through tryptophan-network simulations and ultraviolet spectroscopy
  • sourceinSRC-WIEST-2025AOld theory, new evidence: inhalational anesthetics disrupt a collective quantum state of intraneuronal microtubules to cause unconsciousness
  • sourceinSRC-KERSKENS-2023Reply to Comment on: 'Experimental indications of non-classical brain function' (2022 Journal of Physics Communications 6 105001)
  • claiminORCH-C044Spin dynamics of a naturally occurring tryptophan-cation/superoxide radical pair modulate the microtubule polymerization rate, so that microtubule density depe…
  • claiminORCH-C046LISA Pathfinder's measured relative acceleration noise between two free-falling test masses bounds the Diósi–Penrose regularization length to at least 40.1 ± 0…
  • claiminORCH-C047An independent underground ultra-low-background germanium search, the Majorana Demonstrator, found no evidence of the 1/E spontaneous-collapse radiation signat…
  • claiminORCH-C065E_Δ in the collapse rate is the difference between gravitational self-energies before and after the collapse times an unspecified numeric constant.
  • claiminORCH-C066A semiclassical concept and its mathematical realization by Tilloy and Diósi confirms the heuristic gravity-related collapse rate proposal.
  • claiminORCH-C067Under the Tilloy–Diósi construction, at optimum measurement precision the superposition decays exactly at the rate E_Δ/ħ.
  • claiminORCH-C068The Tilloy–Diósi construction makes the numeric constant in E_Δ unique, with the prefactor const. = 1/2.
  • claiminORCH-C069Penrose and Diósi disagree on how the collapse happens — smoothly, suddenly, or some other way.
  • claiminORCH-C057No candidate molecular target for inhalational anesthesia other than tubulin has a published model that reproduces the Meyer–Overton correlation; it would be f…
  • claiminORCH-C063Including molecular-dynamics-derived structural variability substantially reduces the maximal superradiant enhancement of modelled microtubule tryptophan netwo…
  • evidenceinORCH-E042The four published per-isotope xenon potencies, independently restated
  • evidenceinORCH-E044Radical-pair spin dynamics modelled as a modulator of microtubule polymerization
  • evidenceinORCH-E046LISA Pathfinder bounds the Diósi–Penrose cutoff above any nuclear size — through a channel radiation arguments do not reach
  • evidenceinORCH-E047Majorana Demonstrator: no spontaneous collapse radiation in 37.5 kg-y of enriched germanium
  • evidenceinORCH-E049A second underground null adds weight to the exclusion of the parameter-free DP model
  • evidenceinORCH-E070A second underground null adds weight to the exclusion of the parameter-free DP model — toward ORCH-C030
  • evidenceinORCH-E071The co-originator: the rate carries an unspecified constant and may not be parameter-free — part 1
  • evidenceinORCH-E051A semiclassical construction reproduces Penrose's rate exactly and fixes the constant at 1/2
  • evidenceinORCH-E052Penrose and Diósi agree on the lifetime and disagree on the dynamics
  • evidenceinORCH-E0722026 modelling reports entanglement measures but its lifetimes are radiative, not dephasing, times — part 3
  • evidenceinORCH-E0732026 modelling reports entanglement measures but its lifetimes are radiative, not dephasing, times — part 4
  • evidenceinORCH-E055The modellers' own disclaimer: nothing here implies microtubules must use quantum correlations
  • evidenceinORCH-E056Cross-lab: polymerization raises tubulin quantum yield at 280 nm in porcine tubulin
  • evidenceinORCH-E074Cross-lab: polymerization raises tubulin quantum yield at 280 nm in porcine tubulin — toward ORCH-C013
  • evidenceinORCH-E075The magnitude of the polymerization effect spans nearly an order of magnitude on the scattering correction — part 1
  • evidenceinORCH-E076At tryptophan-selective excitation the polymerization effect reverses sign between treatments — part 1
  • evidenceinORCH-E077At tryptophan-selective excitation the polymerization effect reverses sign between treatments — part 2
  • evidenceinORCH-E078At tryptophan-selective excitation the polymerization effect reverses sign between treatments — part 3
  • evidenceinORCH-E079At tryptophan-selective excitation the polymerization effect reverses sign between treatments — part 4
  • evidenceinORCH-E080Structural variability suppresses the superradiant enhancement the collective reading relies on — part 2
  • evidenceinORCH-E081A quantum-yield measurement becomes 'definitive biophysical evidence' in one step — part 1
  • evidenceinORCH-E082A quantum-yield measurement becomes 'definitive biophysical evidence' in one step — part 2
  • evidenceinORCH-E083A quantum-yield measurement becomes 'definitive biophysical evidence' in one step — part 3
  • evidenceinORCH-E084The proponents' strongest unanswered point: Meyer–Overton from tubulin quantum chemistry — part 1
  • evidenceinORCH-E085The proponents' strongest unanswered point: Meyer–Overton from tubulin quantum chemistry — part 2
  • evidenceinORCH-E086The proponents' strongest unanswered point: Meyer–Overton from tubulin quantum chemistry — part 3
  • evidenceinORCH-E087The zero-quantum-coherence authors' quantitative reply: classical iZQC cannot produce signals of the observed size — part 2
  • evidenceinORCH-E088The zero-quantum-coherence authors' quantitative reply: classical iZQC cannot produce signals of the observed size — part 3
  • evidenceinORCH-E089The zero-quantum-coherence authors' quantitative reply: classical iZQC cannot produce signals of the observed size — part 5
  • evidenceinORCH-E090The zero-quantum-coherence authors' quantitative reply: classical iZQC cannot produce signals of the observed size — part 7
  • researchinORCH-R011Xenon isotope anesthetic potency measured under a controlled external magnetic field
  • researchinORCH-R012Zinc-isotope and ROS-modulation test of the radical-pair route to microtubule density
  • researchinORCH-R013Desk audit: is the dynamics that reproduces ħ/E_Δ already excluded by the bounded channels?
  • researchinORCH-R014Publish, in advance, a 310 K thermal-bath dephasing prediction for the same 1JFF tryptophan geometry
  • researchinORCH-R015Close the scattering ambiguity in the microtubule quantum-yield observable before any potency series
2026-09-11reportresearch pass (claude-opus-5)$3.31

run 2026-09-11-report-orch-or-114458

This run wrote no disposition rows.

2026-09-11editionnew assessment; served by the fallback: claude-fable-5-1 → claude-opus-5 (refusal, bio; 1 fallback iteration)$3.46

run 2026-09-11-edition-orch-or-123904

This run wrote no disposition rows.

2026-09-11draftschema sent as instructions (too large for strict output). Model stamp caveat, added by hand 2026-09-11: this run began before the served-model fix (servedBy, src/pipeline/models.ts) was in place and the draft verb keeps no raw response, so whether the vendor's fallback (claude-opus-5) answered instead of claude-fable-5-1 is unknown; the stamp records the model asked. The research run before it fell back on its first turn.4 blocked · 1 irrelevant$4.65

run 2026-09-11-draft-orch-or-115454

refused, with the reason

  • sourceblockedCited in SRC-SMITH-2021 [p. 1] as showing that specific electron spin resonance signals are observed when xenon acts anesthetically on Drosophila; and at [p. 5…

    Turin, Skoulakis & Horsfield, 'Electron spin changes during general anesthesia in Drosophila', PNAS 111:E3524 (2014) — potentially the most diagnostic item in the spin family, being an in vivo spin measurement under anesthesia rather than a model fit. It was seen only as a reference inside SRC-SMITH-2021 and as a sear…

    reopen if: The full text is retrieved and the reported ESR signals, their oxygen dependence, and the conditions of measurement can be quoted with locators.

  • sourceblockedSRC-WIEST-MGR-2025 states that work from Bandyopadhyay and colleagues 'has demonstrated MT resonances that control membrane spiking activity in living neurons,…

    Singh et al., 'Cytoskeletal filaments deep inside a neuron are not silent', Symmetry 13:821 (2021) — cited by SRC-WIEST-2025A as evidence that microtubule resonances regulate spike timing and span multiple neurons, which bears on ORCH-C022 and on the scale-bridging argument. Not retrieved; a citation inside a commenta…

    reopen if: The MDPI open-access text is retrieved, so the reported resonance-to-spike relationship and its controls can be quoted.

  • sourceblockedSurfaced in search listing only; no content read.

    arXiv:2512.15393, 'Spontaneous wave function collapse from non-local gravitational self-energy' — possibly a live repair attempt on the bounded Diósi–Penrose regularization, which would bear directly on ORCH-C030 and ORCH-C051. The fetch failed (unsupported content type) and no text was seen.

    reopen if: The arXiv abstract and PDF are retrieved and the paper's predictions for spontaneous radiation or momentum diffusion can be quoted.

  • sourceblockedSearch listing only; no content read.

    A radical-pair-mechanism review PDF surfaced in the search returned as base64 and was not parsed; it could not be read and is not cited. Its likely value is background rather than a new observation, since the two primaries it would summarize are now in the record.

    reopen if: The review is retrieved in readable form and contains a primary observation or a bound not already held at ORCH-C101 through ORCH-C104.

  • sourceirrelevantSearch listings and abstracts only; no primary observation identified.

    Several 2026 secondary reviews and self-published syntheses of quantum-consciousness theories (Frontiers reviews, Zenodo and SSRN items) surfaced in search. They are summaries of primaries the ledger already holds and advance no observation, bound or measurement the record lacks; under §3.7 the primaries are preferred.

    reopen if: One of them reports an original measurement, a re-analysis of primary data, or a quantitative acceptance criterion for a coherence test that the Orch OR corpus does not contain.

Change history

Every change to this case, newest first: what changed, why, and who made it — including the AI's role. The log is append-only; a correction is a new entry, never an edit to an old one.

  1. aletheia edition (claude-opus-5-5) · AI-assisted

    Research agenda: ORCH-R008 open → answered (Study ORCH-S001 collected the corpus. ORCH-E035 shows the theory's parameter values are consistent across two decades, that no pass/fail threshold for any direct coherence test is published, and that the carrier changed in 2014. ORCH-E036 shows no unconditional proponent estimate reaches 25 ms. The remaining step, a proponent-endorsed threshold fixed in advance, is carried by the open item ORCH-R014.); ORCH-R010 open → answered (Study ORCH-S002 ran the pre-registered two-covariate re-analysis. ORCH-E037: the frozen identification rule tripped (condition number 302.4, driven by the narrow mass range rather than spin–mass collinearity), so no spin coefficient is evidential. ORCH-E038: a spin step describes the four means far better than a linear mass term, descriptively only. ORCH-E039 and ORCH-E040: the result is unreplicated, its headline values are calculated, and its dispersions are of unstated type. The dataset cannot support the inferential test; a new experiment (ORCH-R011) is what remains.).

    Why: Set by the edition verb (claude-opus-5-5, edition-v14) in run 2026-09-30-edition-orch-or-093603, reading the agenda against the ledger as it stands; the edition edition-2026-09-30-094411 carries the reasoning.

  2. claude-fable-5-1 (operator), answering review note #282 in the founder's session of 2026-09-14 · AI-assisted

    Annotation of the 2026-09-11 intake entry (actor 'aletheia verify (claude-sonnet-5 second reader; drafter claude-fable-5-1)'): the drafter named there is the model asked; the model that served draft run 2026-09-11-draft-orch-or-115454 is unrecorded (that run began before the served-model stamp fix and kept no raw response; the vendor's fallback, claude-opus-5, may have answered). The 21 records the run extracted already carry that caveat in their origin stamps (corrections of 2026-09-11). The history log is append-only, so the earlier line stands as written and this entry qualifies it.

    Why: Review note #282 (GPT-5.6 Sol seat on #281): the history actor should carry the requested-model / unknown-served-model caveat. Answered by this entry rather than an edit, because the changelog is append-only (§3.14).

  3. claude-fable-5-1 (operator), on the founder's direction in session 2026-09-11 (the atomic-statement rule, PR fix/orch-lessons) · AI-assisted

    Corrections applied from the atomic-statement rule (draft-v8), applied to the two statements that landed before it: ORCH-C057.statement — "No candidate molecular target for inhalational anesthesia other than tubulin has a published model that reproduces the Meyer–Overton correlation; it would be false if such a model existed for any receptor, channel or membrane target." → "No candidate molecular target for inhalational anesthesia other than tubulin has a published model that reproduces the Meyer–Overton correlation."; ORCH-C063.statement — "Including molecular-dynamics-derived structural variability substantially reduces the maximal superradiant enhancement of modelled microtubule tryptophan networks relative to an ordered 1JFF reference lattice; it would be false if the disordered ensembles retained the ordered lattice's peak radiative-rate enhancement." → "Including molecular-dynamics-derived structural variability substantially reduces the maximal superradiant enhancement of modelled microtubule tryptophan networks relative to an ordered 1JFF reference lattice.".

    Why: ORCH-C057: §3.2: the statement bundled a falsification condition with the proposition; the condition is removed (what would falsify the claim stays in its evidence and the edition's treatment), as for ORCH-C044/C046/C047 on PR #281 | ORCH-C063: §3.2: the statement bundled a falsification condition with the proposition; the condition is removed (what would falsify the claim stays in its evidence and the edition's treatment), as for ORCH-C044/C046/C047 on PR #281

  4. claude-fable-5-1 (operator), answering the panel's second park on PR #281 in the founder's session of 2026-09-11 · AI-assisted

    Corrections applied from PR #281 panel review, second round (GPT-5.6 Sol seat: compound claim statements): ORCH-C044.statement — "Spin dynamics of a naturally occurring tryptophan-cation/superoxide radical pair modulate the microtubule polymerization rate, so that microtubule density depends on ambient magnetic field strength; it would be false if microtubule assembly showed no dependence on field strength or on radical-pair-relevant redox conditions." → "Spin dynamics of a naturally occurring tryptophan-cation/superoxide radical pair modulate the microtubule polymerization rate, so that microtubule density depends on ambient magnetic field strength."; ORCH-C046.statement — "LISA Pathfinder's measured relative acceleration noise between two free-falling test masses bounds the Diósi–Penrose regularization length to at least 40.1 ± 0.5 fm, a value larger than the size of any nucleus; it would be false if the residual acceleration noise were shown to be larger than reported or if the mapping from collapse force noise to acceleration noise were invalid." → "LISA Pathfinder's measured relative acceleration noise between two free-falling test masses bounds the Diósi–Penrose regularization length to at least 40.1 ± 0.5 fm, a value larger than the size of any nucleus."; ORCH-C047.statement — "An independent underground ultra-low-background germanium search, the Majorana Demonstrator, found no evidence of the 1/E spontaneous-collapse radiation signature in the 19–100 keV range over a 37.5 kg-y enriched exposure; it would be false if a signal excess consistent with the predicted spectral shape were present in that dataset." → "An independent underground ultra-low-background germanium search, the Majorana Demonstrator, found no evidence of the 1/E spontaneous-collapse radiation signature in the 19–100 keV range over a 37.5 kg-y enriched exposure.".

    Why: ORCH-C044: §3.2: the drafted statement bundled the proposition with an independently truth-evaluable falsification condition; the condition is removed from the proposition (the sources, evidence and the edition's treatment already state what would falsify it) | ORCH-C046: §3.2: the drafted statement bundled the proposition with an independently truth-evaluable falsification condition; the condition is removed from the proposition (the sources, evidence and the edition's treatment already state what would falsify it) | ORCH-C047: §3.2: the drafted statement bundled the proposition with an independently truth-evaluable falsification condition; the condition is removed from the proposition (the sources, evidence and the edition's treatment already state what would falsify it)

  5. claude-fable-5-1 (operator), answering the panel's second park on PR #281 in the founder's session of 2026-09-11 · AI-assisted

    Correction applied from PR #281 panel review, second round (GPT-5.6 Sol seat: ORCH-E071 locator broader than the passage quoted): ORCH-E071.exactLocator — "[p. 1] (Sec. I) and [p. 4] (Sec. V), arXiv PDF 2111.04604v2" → "[p. 1] (Sec. I), arXiv PDF 2111.04604v2".

    Why: ORCH-E071: after the split this record quotes only the rate relation with its "unspecified numeric constant", which the paper states on p. 1 (Sec. I); the p. 4 (Sec. V) half of the locator belonged to the other half of the pre-split record and is dropped

  6. claude-fable-5-1 (operator), answering the panel's park on PR #281 in the founder's session of 2026-09-11 · AI-assisted

    Corrections applied from PR #281 panel review (GPT-5.6 Sol seat: provenance stamps), by direct edit because these two origin blocks carry YAML anchors the correction writer cannot splice: ORCH-E049.origin.extractedBy — "claude-fable-5-1" → "claude-fable-5-1 asked; served model unrecorded for this run — the vendor's fallback (claude-opus-5) may have answered; see proposals/2026-09-11-draft-orch-or-115454/run.yaml"; ORCH-E056.origin.extractedBy — "claude-fable-5-1" → "claude-fable-5-1 asked; served model unrecorded for this run — the vendor's fallback (claude-opus-5) may have answered; see proposals/2026-09-11-draft-orch-or-115454/run.yaml"

    Why: ORCH-E049: the run that extracted this record began before the served-model stamp fix and kept no raw response, so the model that answered is unknown; the stamp says so instead of asserting the model asked | ORCH-E056: the run that extracted this record began before the served-model stamp fix and kept no raw response, so the model that answered is unknown; the stamp says so instead of asserting the model asked

  7. claude-fable-5-1 (operator), answering the panel's park on PR #281 in the founder's session of 2026-09-11 · AI-assisted

    Corrections applied from PR #281 panel review (GPT-5.6 Sol seat: provenance stamps asserted the model asked while the run's served model is unknown): ORCH-C044.origin — {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"} → {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1 asked; served model unrecorded for this run — the vendor's fallback (claude-opus-5) may have answered; see proposals/2026-09-11-draft-orch-or-115454/run.yaml","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"}; ORCH-C046.origin — {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"} → {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1 asked; served model unrecorded for this run — the vendor's fallback (claude-opus-5) may have answered; see proposals/2026-09-11-draft-orch-or-115454/run.yaml","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"}; ORCH-C047.origin — {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"} → {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1 asked; served model unrecorded for this run — the vendor's fallback (claude-opus-5) may have answered; see proposals/2026-09-11-draft-orch-or-115454/run.yaml","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"}; ORCH-C057.origin — {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"} → {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1 asked; served model unrecorded for this run — the vendor's fallback (claude-opus-5) may have answered; see proposals/2026-09-11-draft-orch-or-115454/run.yaml","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"}; ORCH-C063.origin — {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"} → {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1 asked; served model unrecorded for this run — the vendor's fallback (claude-opus-5) may have answered; see proposals/2026-09-11-draft-orch-or-115454/run.yaml","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"}; ORCH-E042.origin — {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"} → {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1 asked; served model unrecorded for this run — the vendor's fallback (claude-opus-5) may have answered; see proposals/2026-09-11-draft-orch-or-115454/run.yaml","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"}; ORCH-E044.origin — {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"} → {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1 asked; served model unrecorded for this run — the vendor's fallback (claude-opus-5) may have answered; see proposals/2026-09-11-draft-orch-or-115454/run.yaml","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"}; ORCH-E046.origin — {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"} → {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1 asked; served model unrecorded for this run — the vendor's fallback (claude-opus-5) may have answered; see proposals/2026-09-11-draft-orch-or-115454/run.yaml","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"}; ORCH-E047.origin — {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"} → {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1 asked; served model unrecorded for this run — the vendor's fallback (claude-opus-5) may have answered; see proposals/2026-09-11-draft-orch-or-115454/run.yaml","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"}; ORCH-E070.origin — {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"} → {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1 asked; served model unrecorded for this run — the vendor's fallback (claude-opus-5) may have answered; see proposals/2026-09-11-draft-orch-or-115454/run.yaml","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"}; ORCH-E051.origin — {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"} → {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1 asked; served model unrecorded for this run — the vendor's fallback (claude-opus-5) may have answered; see proposals/2026-09-11-draft-orch-or-115454/run.yaml","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"}; ORCH-E052.origin — {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"} → {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1 asked; served model unrecorded for this run — the vendor's fallback (claude-opus-5) may have answered; see proposals/2026-09-11-draft-orch-or-115454/run.yaml","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"}; ORCH-E055.origin — {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"} → {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1 asked; served model unrecorded for this run — the vendor's fallback (claude-opus-5) may have answered; see proposals/2026-09-11-draft-orch-or-115454/run.yaml","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"}; ORCH-E074.origin — {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"} → {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1 asked; served model unrecorded for this run — the vendor's fallback (claude-opus-5) may have answered; see proposals/2026-09-11-draft-orch-or-115454/run.yaml","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"}; ORCH-R011.origin — {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"} → {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1 asked; served model unrecorded for this run — the vendor's fallback (claude-opus-5) may have answered; see proposals/2026-09-11-draft-orch-or-115454/run.yaml","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"}; ORCH-R012.origin — {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"} → {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1 asked; served model unrecorded for this run — the vendor's fallback (claude-opus-5) may have answered; see proposals/2026-09-11-draft-orch-or-115454/run.yaml","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"}; ORCH-R013.origin — {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"} → {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1 asked; served model unrecorded for this run — the vendor's fallback (claude-opus-5) may have answered; see proposals/2026-09-11-draft-orch-or-115454/run.yaml","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"}; ORCH-R014.origin — {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"} → {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1 asked; served model unrecorded for this run — the vendor's fallback (claude-opus-5) may have answered; see proposals/2026-09-11-draft-orch-or-115454/run.yaml","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"}; ORCH-R015.origin — {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"} → {"ref":"report 2026-09-11-report-orch-or-114458","extractedBy":"claude-fable-5-1 asked; served model unrecorded for this run — the vendor's fallback (claude-opus-5) may have answered; see proposals/2026-09-11-draft-orch-or-115454/run.yaml","runId":"2026-09-11-draft-orch-or-115454","date":"2026-09-11"}.

    Why: ORCH-C044: the run that extracted this record began before the served-model stamp fix and kept no raw response, so the model that answered is unknown; the stamp says so instead of asserting the model asked | ORCH-C046: the run that extracted this record began before the served-model stamp fix and kept no raw response, so the model that answered is unknown; the stamp says so instead of asserting the model asked | ORCH-C047: the run that extracted this record began before the served-model stamp fix and kept no raw response, so the model that answered is unknown; the stamp says so instead of asserting the model asked | ORCH-C057: the run that extracted this record began before the served-model stamp fix and kept no raw response, so the model that answered is unknown; the stamp says so instead of asserting the model asked | ORCH-C063: the run that extracted this record began before the served-model stamp fix and kept no raw response, so the model that answered is unknown; the stamp says so instead of asserting the model asked | ORCH-E042: the run that extracted this record began before the served-model stamp fix and kept no raw response, so the model that answered is unknown; the stamp says so instead of asserting the model asked | ORCH-E044: the run that extracted this record began before the served-model stamp fix and kept no raw response, so the model that answered is unknown; the stamp says so instead of asserting the model asked | ORCH-E046: the run that extracted this record began before the served-model stamp fix and kept no raw response, so the model that answered is unknown; the stamp says so instead of asserting the model asked | ORCH-E047: the run that extracted this record began before the served-model stamp fix and kept no raw response, so the model that answered is unknown; the stamp says so instead of asserting the model asked | ORCH-E070: the run that extracted this record began before the served-model stamp fix and kept no raw response, so the model that answered is unknown; the stamp says so instead of asserting the model asked | ORCH-E051: the run that extracted this record began before the served-model stamp fix and kept no raw response, so the model that answered is unknown; the stamp says so instead of asserting the model asked | ORCH-E052: the run that extracted this record began before the served-model stamp fix and kept no raw response, so the model that answered is unknown; the stamp says so instead of asserting the model asked | ORCH-E055: the run that extracted this record began before the served-model stamp fix and kept no raw response, so the model that answered is unknown; the stamp says so instead of asserting the model asked | ORCH-E074: the run that extracted this record began before the served-model stamp fix and kept no raw response, so the model that answered is unknown; the stamp says so instead of asserting the model asked | ORCH-R011: the run that extracted this record began before the served-model stamp fix and kept no raw response, so the model that answered is unknown; the stamp says so instead of asserting the model asked | ORCH-R012: the run that extracted this record began before the served-model stamp fix and kept no raw response, so the model that answered is unknown; the stamp says so instead of asserting the model asked | ORCH-R013: the run that extracted this record began before the served-model stamp fix and kept no raw response, so the model that answered is unknown; the stamp says so instead of asserting the model asked | ORCH-R014: the run that extracted this record began before the served-model stamp fix and kept no raw response, so the model that answered is unknown; the stamp says so instead of asserting the model asked | ORCH-R015: the run that extracted this record began before the served-model stamp fix and kept no raw response, so the model that answered is unknown; the stamp says so instead of asserting the model asked

  8. claude-fable-5-1 (operator), answering the panel's park on PR #281 in the founder's session of 2026-09-11 · AI-assisted

    Corrections applied from PR #281 panel review (GPT-5.6 Sol seat: stale record locators in ORCH-R011 to ORCH-R015): ORCH-R011.summary — "Randomized, blinded loss-of-righting-reflex dose–response in mice for spin-0 and spin-bearing xenon isotopes, run inside a controllable-field enclosure with arms at geomagnetic field, at a field an order of magnitude higher, and at a field about two orders above geomagnetic (i.e. above the upper limit of the field range recorded at ORCH-C102), with field strength logged and the predicted isotope ordering preregistered for each arm. Cheapest decisive discriminator in the case: a field dependence of the isotope effect would support the radical-pair mechanism at ORCH-C101 and is orthogonal to Orch OR, since no microtubule coherence is required; no field dependence at fields well above the model's stated agreement range would falsify the published fit and return the xenon anomaly to unexplained, further reducing what ORCH-C025 can do for ORCH-C002. My searches did not locate such an in vivo experiment; that is an unresolved retrieval question rather than a finding of absence." → "Randomized, blinded loss-of-righting-reflex dose–response in mice for spin-0 and spin-bearing xenon isotopes, run inside a controllable-field enclosure with arms at geomagnetic field, at a field an order of magnitude higher, and at a field about two orders above geomagnetic (i.e. above the upper limit of the field range over which the published radical-pair fit holds (SRC-SMITH-2021)), with field strength logged and the predicted isotope ordering preregistered for each arm. Cheapest decisive discriminator in the case: a field dependence of the isotope effect would support the radical-pair mechanism proposed in SRC-SMITH-2021 and is orthogonal to Orch OR, since no microtubule coherence is required; no field dependence at fields well above the model's stated agreement range would falsify the published fit and return the xenon anomaly to unexplained, further reducing what ORCH-C025 can do for ORCH-C002. My searches did not locate such an in vivo experiment; that is an unresolved retrieval question rather than a finding of absence."; ORCH-R012.summary — "Blinded in vitro and cell-culture comparison of microtubule density and polymerization kinetics under matched chemical concentrations of nuclear-spin-bearing 67Zn versus zero-spin zinc isotopes, crossed with pharmacological modulation of reactive-oxygen-species removal and with field-shielded versus geomagnetic conditions. An isotope difference at matched chemistry, and a field dependence that tracks it, would support the mechanism at ORCH-C103 and ORCH-C104 and give the 'same-direction stabilizer and destabilizer' result at ORCH-C024 a redox-mediated reading with its own predictions; a null on both would leave general cellular disruption as the better account of ORCH-C024 and would remove this rival from the anesthesia arm." → "Blinded in vitro and cell-culture comparison of microtubule density and polymerization kinetics under matched chemical concentrations of nuclear-spin-bearing 67Zn versus zero-spin zinc isotopes, crossed with pharmacological modulation of reactive-oxygen-species removal and with field-shielded versus geomagnetic conditions. An isotope difference at matched chemistry, and a field dependence that tracks it, would support the mechanism at ORCH-C044 and its zinc-isotope prediction (SRC-ZADEHHAGHIGHI-2022) and give the 'same-direction stabilizer and destabilizer' result at ORCH-C024 a redox-mediated reading with its own predictions; a null on both would leave general cellular disruption as the better account of ORCH-C024 and would remove this rival from the anesthesia arm."; ORCH-R013.summary — "From the published literature alone, compute whether the least-decoherence semiclassical construction recorded at ORCH-C109 — which reproduces the conjectured rate with the prefactor fixed — predicts spontaneous radiation and momentum diffusion at levels reached by the Gran Sasso search (ORCH-C031), the independent germanium search (ORCH-C106, ORCH-C107), and the LISA Pathfinder acceleration-noise bound (ORCH-C105); and whether its regularization scale can be pushed above the bounded values without altering the rate. If it is already excluded, the physics conjunct is refuted in every dynamics so far proposed for it and the escape at ORCH-E019 buys nothing; if it survives, the surviving parameter window should be stated numerically, which is what a decisive mesoscopic test (ORCH-R007) would then have to reach." → "From the published literature alone, compute whether the least-decoherence semiclassical construction recorded at ORCH-C066 to ORCH-C068 — which reproduces the conjectured rate with the prefactor fixed — predicts spontaneous radiation and momentum diffusion at levels reached by the Gran Sasso search (ORCH-C031), the independent germanium search (ORCH-C047; its Diósi–Penrose reading at ORCH-E049), and the LISA Pathfinder acceleration-noise bound (ORCH-C046); and whether its regularization scale can be pushed above the bounded values without altering the rate. If it is already excluded, the physics conjunct is refuted in every dynamics so far proposed for it and the escape at ORCH-E019 buys nothing; if it survives, the surviving parameter window should be stated numerically, which is what a decisive mesoscopic test (ORCH-R007) would then have to reach."; ORCH-R014.summary — "Ask the group that already has the machinery to extend its radiative-loss Lindblad treatment (ORCH-C112) to include solvent, ionic and vibrational baths at 310 K, and to publish, before any spectroscopy is run, the predicted dephasing time, coherence length and participating degrees of freedom, together with the value of a named density-matrix witness that would count as pass and the value that would count as fail. This is the pass/fail threshold the case's own audit found missing from the corpus (ORCH-E035, ORCH-E036), and it is the prior question to the expensive experiment at ORCH-R001: without it, any measured number can be absorbed either way." → "Ask the group that already has the machinery to extend its radiative-loss Lindblad treatment (recorded at ORCH-E072 and ORCH-E073, SRC-GASSAB-2026) to include solvent, ionic and vibrational baths at 310 K, and to publish, before any spectroscopy is run, the predicted dephasing time, coherence length and participating degrees of freedom, together with the value of a named density-matrix witness that would count as pass and the value that would count as fail. This is the pass/fail threshold the case's own audit found missing from the corpus (ORCH-E035, ORCH-E036), and it is the prior question to the expensive experiment at ORCH-R001: without it, any measured number can be absorbed either way."; ORCH-R015.summary — "Re-measure the tubulin-dimer versus taxol-stabilized-microtubule quantum yield in cuvette and integrating-sphere geometries, with independent turbidity measurement, inner-filter controls, matched buffers and a measured polymer fraction, at both 280 and 295 nm excitation. The target is to replace the bounded estimate at ORCH-C119 with a single value and to fix the sign at 295 nm (ORCH-C120), and to repeat the non-anesthetic additive control of ORCH-C121 alongside a convulsant and a hydrophobic nonimmobilizer. Until this is closed, an anesthetic potency series against this observable (ORCH-R002, ORCH-R009) risks measuring the scattering correction rather than the physics." → "Re-measure the tubulin-dimer versus taxol-stabilized-microtubule quantum yield in cuvette and integrating-sphere geometries, with independent turbidity measurement, inner-filter controls, matched buffers and a measured polymer fraction, at both 280 and 295 nm excitation. The target is to replace the bounded estimate recorded at ORCH-E075 with a single value and to fix the sign at 295 nm (ORCH-E076 through ORCH-E079), and to repeat the non-anesthetic additive control reported in SRC-GASSAB-2026A alongside a convulsant and a hydrophobic nonimmobilizer. Until this is closed, an anesthetic potency series against this observable (ORCH-R002, ORCH-R009) risks measuring the scattering correction rather than the physics.".

    Why: ORCH-R011: the summary named claim ids (ORCH-C101 to ORCH-C121) from the drafter's provisional numbering that never entered the ledger; re-pointed to the records that survived verification (accepted claims, their evidence, or the source itself) | ORCH-R012: the summary named claim ids (ORCH-C101 to ORCH-C121) from the drafter's provisional numbering that never entered the ledger; re-pointed to the records that survived verification (accepted claims, their evidence, or the source itself) | ORCH-R013: the summary named claim ids (ORCH-C101 to ORCH-C121) from the drafter's provisional numbering that never entered the ledger; re-pointed to the records that survived verification (accepted claims, their evidence, or the source itself) | ORCH-R014: the summary named claim ids (ORCH-C101 to ORCH-C121) from the drafter's provisional numbering that never entered the ledger; re-pointed to the records that survived verification (accepted claims, their evidence, or the source itself) | ORCH-R015: the summary named claim ids (ORCH-C101 to ORCH-C121) from the drafter's provisional numbering that never entered the ledger; re-pointed to the records that survived verification (accepted claims, their evidence, or the source itself)

  9. aletheia verify (claude-sonnet-5 second reader; drafter claude-fable-5-1) · AI-assisted

    Intake from report proposals/2026-09-11-report-orch-or-114458/report.md: 9 source(s), 30 evidence record(s), 10 claim(s), 5 research item(s) verified and added (proposal 2026-09-11-draft-orch-or-115454, verification 2026-09-11-verify-orch-or-120610); 38 candidate(s) rejected with reasons in dispositions.yaml.

    Why: Admitted after verification: SRC-SMITH-2021, SRC-ZADEHHAGHIGHI-2022, SRC-HELOU-2017, SRC-ARNQUIST-2023, SRC-DISI-2022, SRC-GASSAB-2026, SRC-GASSAB-2026A, SRC-WIEST-2025A, SRC-KERSKENS-2023, ORCH-E042, ORCH-E044, ORCH-E046, ORCH-E047, ORCH-E049, ORCH-E070, ORCH-E071, ORCH-E051, ORCH-E052, ORCH-E072, ORCH-E073, ORCH-E055, ORCH-E056, ORCH-E074, ORCH-E075, ORCH-E076, ORCH-E077, ORCH-E078, ORCH-E079, ORCH-E080, ORCH-E081, ORCH-E082, ORCH-E083, ORCH-E084, ORCH-E085, ORCH-E086, ORCH-E087, ORCH-E088, ORCH-E089, ORCH-E090, ORCH-C044, ORCH-C046, ORCH-C047, ORCH-C065, ORCH-C066, ORCH-C067, ORCH-C068, ORCH-C069, ORCH-C057, ORCH-C063, ORCH-R011, ORCH-R012, ORCH-R013, ORCH-R014, ORCH-R015; everything else proposed was refused or blocked with a reason in dispositions.yaml. The drafter's rationale for the proposal, written before verification and describing what it proposed: This pass adds four independent lines the ledger did not hold. (1) A published radical-pair account of the xenon isotope anomaly — quantum, but with no microtubule, no coherent tubulin lattice and no collapse — enters as its own competing mechanism claim linked as an alternative to ORCH-C002 and ORCH-C004 rather than folded into either, together with its field-strength condition, which is the cheapest decisive test in the case. (2) The same group's radical-pair route to magnetic-field effects on microtubule density enters as a mechanistic rival to the 'general cellular disruption' reading of the same-direction stabilizer/destabilizer anomaly, with a zinc-isotope prediction. (3) The collapse-physics rung gains two further non-interferometric constraints (LISA Pathfinder's mechanical-noise channel, which the radiation-specific escape of ORCH-E019 does not reach, and an independent underground germanium search) plus the co-originator's own primary statement of what the rate conjecture is and is not. (4) The 2026 microtubule photophysics literature is recorded honestly: entanglement-flavoured observables computed in a Markovian radiative-loss model with no thermal bath, disorder suppressing long-range coherence, and a cross-lab quantum-yield experiment whose polymerization magnitude spans nearly an order of magnitude on the scattering correction, with free L-tryptophan quenching the same observable — a second instance of the ORCH-E003 specificity problem. The Kerskens–Pérez reply, which the report could not reach but whose text was supplied here, is added so that ORCH-C026 is no longer comment-without-reply. Candidates whose text was not shown (Turin 2014, the Bandyopadhyay-group Symmetry paper, arXiv:2512.15393, the radical-pair review PDF) stay out with routes. No edition candidate: nothing here removes the absence on which the verdict rests, and the new rival account is on the record for a later telling to weigh.

  10. claude-fable-5-1 (operator), on the founder's direction in session 2026-09-09 · AI-assisted

    The editorial conjecture file retired (ORCH-J001, by "Eugene (ejhong)", dated 2026-08-23, status open). The card leaves the case page. Its propositions are the ledger's: ORCH-C001 (each conscious moment is an orchestrated objective reduction in microtubules) and ORCH-C002 (anesthetics act substantially by disrupting a quantum-coherent microtubule state); its decisive tests are ORCH-R001, ORCH-R002, ORCH-R004 and ORCH-R007; its rationale is the founder's essay "Orch OR Ascending", a founding input of this case (ORCH-IN002), which the file said states the reasons in full. Banked here, as the file recorded them, are the founder's stake and the tests he named. Statement: "Orch OR is substantially correct: consciousness involves quantum processes in neuronal microtubules terminated by Penrose objective reduction, and anesthesia works substantially by dampening those quantum effects." Confidence: "High — stated as founder's intuition, explicitly not derived from the evidence ledger, which currently grades the corresponding claims (ORCH-C001, ORCH-C002) unresolved and weakly supported." Predicted findings: (1) "A validated coherence witness in microtubules under physiological conditions showing quantum effects persisting far beyond naive warm-wet-noisy decoherence estimates." (2) "Anesthetics dampen microtubule quantum observables (excitation diffusion, delayed luminescence, resonance patterns) in proportion to clinical potency, while structurally similar non-anesthetics do not — replicated under preregistered protocols." (3) "The microtubule-stabilizer anesthesia effects (epothilone) and the xenon nuclear-spin isotope potency split survive independent preregistered replication." (4) "BEC or levitated-superposition experiments detect state reduction on the τ ≈ ℏ/E_G timescale Penrose's formula predicts, as superposed masses scale up." What would prove it wrong: (1) "A validated physiological coherence measurement (ORCH-R001-class) finds decoherence times orders of magnitude too short, with sensitivity demonstrated on positive controls — the microtubule substrate claim should then be marked refuted." (2) "Preregistered anesthetic potency series show no correlation between clinical potency and any microtubule quantum observable, or the epothilone and xenon-isotope anomalies fail replication — the anesthesia pathway claim should then be dropped, not relocated." (3) "Fuentes-program superposition experiments reach the mass scales where τ ≈ ℏ/E_G predicts collapse and observe standard unitary quantum mechanics — gravitational OR is then wrong and the theory's physical foundation fails."

    Why: Founder direction (2026-09-09, in session): a bet with a stated confidence and a disconfirmer is a claim, graded honestly against the evidence, not a registry beside the ledger; the essays are founding inputs and their propositions enter as claims. What the founder believed ahead of the evidence, and what he staked it on, belongs in the append-only changelog (AGENTS.md §7, amendment of 2026-09-01). The conjecture carried no evidential weight, so no standing changes; the file's text remains in git history.

  11. scripts/migrations/2026-09-09-hash-locators.ts (mechanical re-stamp) · AI-assisted

    Ledger hash redefined to exclude source locators (url, archivedUrl); 1 edition/assessment basis stamp(s) recomputed mechanically, content unchanged: edition-2026-09-08-migration.yaml.

    Why: Founder decision 2026-09-09: a locator says where a record is read, not what it says; a by-hand URL addition had set a fresh panel aside and made an edition due that could only re-adopt.

  12. Cursor Grok 4.6 (Cursor cloud agent), directed by the founder; runId title-punctuation-2026-09-08; promptVersion none (founder-directed session, not a pipeline prompt) · AI-assisted

    Case title punctuation restored after the pre-overhaul publication restore: "Collapse, Not Computation?" is again "Collapse, Not Computation". The trailing question mark is dropped so the published name is a statement. Subtitle and framing questions are unchanged.

    Why: Founder direction (first given 2026-09-06, reaffirmed 2026-09-08 after the restore): case names should not carry question marks. The interrogative belongs in the subtitle and the research questions, not on the card and in the header.

  13. scripts/migrations/2026-09-08-editions.ts (mechanical), founder-directed session · AI-assisted

    Migrated to editions (docs/AUTOMATION.md, build step 2). The article, the 19-claim featured set, and the displayed judgment now live in edition edition-2026-09-08-migration, which adopts the migration assessment 2026-09-08-migration-orch-or. Claim records carry propositions and anchors only; evaluative fields moved into the adopted assessment's per-claim treatment; the dossier header, research priority, and component verdicts moved from case.yaml into the assessment; claims-catalog.yaml merged into claims.yaml (tiers ended). No judgment changed. Where a claim record's hand-written credibility differed from the draft overlay's verdict, the record's value (what claim pages displayed) was carried: ORCH-C025: record provisionally_supported, draft weakly_supported.

    Why: One reader-facing unit that binds verdict, selection, and article, so they cannot drift apart; one home for judgment; featuring becomes a recorded decision.

  14. Aletheia editorial AI (Claude Opus 5, Cursor cloud agent) · AI-assisted

    Study ORCH-S002 collected: four rows, five findings, criteria hash 4fa792b878fa re-verified unchanged from the freeze. The pre-committed analysis code is published beside the study file. Four findings enter the ledger. ORCH-E037 (qualifies C025): the frozen identification-failure rule trips, so no spin p-value is reported as evidential — but the variance inflation factor is 3.25, below the threshold of 5, and what actually fails is the scaled condition number at 302.4, whose source is intercept-versus-mass rather than spin-versus-mass, because atomic mass varies by under 4 per cent. ORCH-E038 (supports C025, weak): descriptively, a spin step fits the four means far better than a linear mass trend, since the within-spin-class mass slopes are small and of opposite sign while the one-mass-unit step at the class boundary is 29 per cent atmosphere. ORCH-E039 (undermines C025): no replication, no extension, and no pre-2018 isotope-resolved report exists across 37 forward citations, 24 backward references, three preprint servers and two registries. ORCH-E040 (undermines C025): the reported dispersions are of unstated type and the headline xenon-alone values are calculated rather than measured.

    Why: Executing ORCH-R010 at desk cost, per the pre-registration that merged in PR #152. The freeze's own recorded expectation is scored in the findings as right on the verdict and wrong on the mechanism: it predicted spin/mass collinearity would defeat the regression, and the defeating collinearity turned out to be intercept/mass instead. Because the ill-conditioning has that source, the study can name the decisive missing experiment precisely — a spin-zero isotope lighter than 129, such as 124Xe or 126Xe, which the current design lacks entirely and which would separate "lighter is more potent" from "spinless is more potent" directly.

  15. Aletheia editorial AI (Claude Opus 5, Cursor cloud agent) · AI-assisted

    Review of the Bench's ORCH-S002 draft before it left the working tree. The three citations the drafter wrote into the hash-frozen criteria were independently resolved against PubMed and Crossref, because a drafting model naming specific papers inside criteria that can never afterwards be edited is the one place a fabricated reference would become permanent. All three verify exactly: Li N, Lu D, Yang L, Tao H, Xu Y, Wang C, Fu L, Liu H, Chummum Y, Zhang S, Anesthesiology 2018;129(2):271-277 (PMID 29642079, doi 10.1097/ALN.0000000000002226); Turin, Skoulakis & Horsfield, PNAS 2014;111(34):E3524-E3533 (doi 10.1073/pnas.1404387111); and Turin & Skoulakis, Methods Enzymol 2018;603:115-128 (doi 10.1016/bs.mie.2018.01.020). Bibliographic details only were checked — no potency number was read, and no query from the frozen search protocol was run, both of which belong to collection after this freeze survives the gates. The study title was corrected to drop a "STUDY FREEZE:" prefix the drafter had echoed from its own instructions into a field that renders as the page heading; the title sits outside the criteria hash, which was re-verified unchanged at 4fa792b878fa.

    Why: The Bench drafts freezes with a model, and the criteria hash makes those criteria permanent at merge. Verifying the named citations before the hash becomes binding is the cheapest point at which a fabricated reference can still be caught, and AGENTS.md 3.15 forbids one absolutely.

  16. Bench freeze drafter (bench-freeze-v1) · AI-assisted

    Study ORCH-S002 pre-registered by the Bench: "Xenon isotope potency regressed on nuclear spin and atomic mass, from published per-isotope values" (executes new research item ORCH-R010), drafted from the panel-advanced agenda proposal "Does xenon isotope potency track nuclear spin or atomic mass? A regression on the published isotope data" (2026-08-31-agenda-33432478115/orch-or/1). Zero rows by design; criteria hash-frozen; collection follows the two-PR discipline after this freeze survives the gates.

    Why: Bench v2 (docs/AUTOMATION.md): proposals scored high-gain by four of five seats with no constitutional concern auto-draft their freeze; the panel judges the protocol blind to any data.

  17. anthropic/claude-opus-5 (scheduled maintenance run, aletheia-editorial-audit-v1); publishes only through the constitutional gate's multi-model concurrence, per AGENTS.md §3.15 · AI-assisted

    Corrected the shortfall figure in the decoherence section: the proponents' own revised 10⁻⁵–10⁻⁴ s estimate falls two to three orders of magnitude short of the theory's 25 ms exemplar, not one to two, per the tabulation of both sides' published numbers. Updated the closing paragraph, which described the desk audit of what Orch OR commits to numerically as a study still awaiting funding on the agenda; that audit has since been carried out. Rewrote research item ORCH-R008 accordingly, recording what the tabulation found — consistent parameter values, no acceptance criterion, a carrier changed in the 2014 revision — and narrowing the remaining work to obtaining a proponent-endorsed pass/fail threshold in advance of the coherence and potency measurements.

    Why: The article and agenda proposed as future work a corpus audit that the records now report as done, and stated a numerical shortfall that the newer tabulation contradicts.

  18. Aletheia editorial AI (Kimi K3, Cursor cloud agent) · AI-assisted

    Study ORCH-S001 collected: nineteen (publication, parameter) rows tabulated from the twelve candidates frozen this morning (hash-verified unchanged), with two aggregate findings entering the ledger (ORCH-E035, ORCH-E036) via the study's workpaper source. The corpus's numbers are internally consistent and independently reproducible, but the physical carrier changed between formulations, no unconditional proponent decoherence estimate meets the theory's own requirement, and the acceptance criteria a decisive test would need are nowhere committed.

    Why: Collection phase of the case's first pre-registered study, per the two-PR discipline; executes ORCH-R008.

  19. Aletheia editorial AI (Kimi K3, Cursor cloud agent) · AI-assisted

    Study ORCH-S001 pre-registered (freeze only, zero rows): the quantitative commitments of the Orch OR literature, criteria hash ca78e87d14cc. One row per (publication, parameter) pair, proponent and critique under identical rules, with the stated-as column frozen in advance — whether each number is a requirement, an estimate, an invoked measurement, an assumption, or a prediction, taken from the paper's own wording. Twelve known candidates pre-committed with dispositions. Collection follows in a separate PR.

    Why: ORCH-R008 names the defect that makes every other item on the case's agenda harder to interpret — without a published numerical specification, no measurement can count as pass or fail, and each new photophysics result gets absorbed rather than scored. The corpus's numbers belong in one table, assembled under frozen criteria so the table cannot be tuned to flatter either the theory or its critics; a documented absence of a specification is itself the most useful finding available at desk cost.

  20. anthropic/claude-opus-5 (scheduled maintenance run, aletheia-editorial-audit-v1); publishes only through the constitutional gate's multi-model concurrence, per AGENTS.md §3.15 · AI-assisted

    Corrected the closing section on the conventional account, which stated that its only weakness was explanatory rather than mechanistic. The record from Eger and colleagues reports that no ligand- or voltage-gated channel, and no combination of them, has been found sufficient to mediate immobility from inhaled anesthetics, so the mechanistic account is itself incomplete for those agents. The sentence now says so while leaving the explanatory point intact.

    Why: As written, the article presented the classical mechanism of anesthesia as settled in a way the evidence records contradict, which would mislead a reader about where the rival account is and is not complete.

  21. Fable (AI agent), directed by Eugene · AI-assisted

    Correction recorded during the same intake: the taxane/anesthesia clinical observation (Linganna et al. 2015) is frequently paraphrased as showing that patients on microtubule stabilizers required more anesthetic. The published endpoint is a hemodynamic surrogate — a greater blood-pressure change on incision — not a measured difference in dose requirement. ORCH-E032 records the study with that correction and the retrospective's confounds attached.

    Why: Observation before interpretation: a paraphrase that upgrades a surrogate endpoint into a dose measurement must not enter the ledger uncorrected, whichever side it favors.

  22. Fable (AI agent), directed by Eugene · AI-assisted

    Intake prompted by private correspondence from a proponent-side researcher (runId orch-pointers-2026-08-24). The correspondence is not published, not quoted, and is not a source: it was used only as a pointer to literature, and every citation below was located and verified independently before entry. Added five sources (SRC-EGER-2008, SRC-JURD-2007, SRC-CARHART-HARRIS-2012, SRC-LINGANNA-2015, SRC-HAMEROFF-2018-EDITORIAL) and five evidence records (ORCH-E030–E034). Sharpened ORCH-C004's credibility summary with two qualifications the new records support: the classical literature's own flagship review reports that no channel or combination suffices even for the immobility endpoint it studied, and the β3(N265M) receptor result is endpoint-specific — immobility abolished, the consciousness-adjacent endpoint reduced but not abolished. Added research opportunity ORCH-R009: test whether a microtubule optical observable's anesthetic dose-response tracks clinical potency across a chemically diverse panel — a Meyer–Overton correlation for the structure the theory nominates.

    Why: Fairness and negative-evidence rules: a proponent's strongest citations belong in the ledger, and so do the corrections that verification produced against them.

  23. Fable (AI agent), directed by Eugene · AI-assisted

    Fifth independent judge added to the cross-model check: Kimi K3 (Moonshot's open-weights model, served via Venice) re-assessed the case blind under the same protocol as the four frontier judges. Its case verdict and per-claim verdicts appear on the concurrence panel and claim pages alongside the others; concurrence statistics recomputed over five judges.

    Why: An open-weights judge from a different training lineage strengthens the independence of the check — agreement across five vendors, including one outside the Western frontier labs, is harder to attribute to shared disposition.

  24. Fable (AI agent), directed by Eugene · AI-assisted

    Cross-model check published (four overlays, role check): the 2026-08-23 multi-model experiment's independent judge runs — GPT-5.1 (OpenAI), Gemini 3.1 Pro (Google), Grok 4.6 (xAI), and Opus 5 (Anthropic), each re-assessing the case blind to all prior assessments — moved from proposals into this case's assessments directory as append-only check overlays. Result now visible on the case page: unanimous case verdict (unresolved) matching the displayed draft; claim-level agreement 15/18 exact, 3 within one step, 0 split. Check runs never display as the case narrative; the concurrence panel summarizes them, and claim pages list each model's per-claim verdict. Note: these runs predate the ORCH-C003/C004 split and assess the pre-split C003 wording; the advocate runs from the same experiment remain in proposals/assessment-experiments/ as experiment artifacts, not checks.

    Why: A verdict corroborated blind by four independent vendors is a property of the evidence file, not one model's disposition — that fact belongs on the public page, not in a proposals folder.

  25. Eugene (conjecture); recorded by Fable (AI agent) · AI-assisted

    Editorial conjecture recorded (ORCH-J001, conjectures.yaml): the founder's on-the-record bet that Orch OR is substantially correct, with the rationale imported from his "Orch OR Ascending" essay and the decisive tests staked on two tracks — anesthetic dampening of microtubule quantum observables in proportion to clinical potency (ORCH-R001/R002/R004), and the Fuentes program's direct tests of Penrose objective reduction at the ℏ/E_G timescale (ORCH-R007). Conjectures carry no evidential weight and do not touch any claim's credibility.

    Why: A truth-seeking system must be able to demonstrate that its founder's best intuition was wrong; recording the bet with disconfirmers is how.

  26. Eugene (direction); applied by Fable (AI agent) · AI-assisted

    ORCH-C003 split. The former claim graded "well supported" a statement whose own summary conceded its universal quantifier was not established ("fully accounted for... without any quantum-coherent contribution"). C003 now carries only the narrow, well-evidenced core (classical pharmacology substantially accounts for anesthetic unconsciousness, with causal evidence for intravenous agents), and the exhaustive version is a new claim ORCH-C004, graded unresolved. The error was flagged independently by an external model review and by the 2026-08-23 multi-model assessment experiment, in which both role-conditioned advocate runs and two of five judges declined to endorse the compound wording.

    Why: Atomic-claims rule: a receptor-level result must not silently underwrite a universally quantified negative about consciousness.

  27. Eugene (direction); applied by Fable (AI agent) · AI-assisted

    Site v1.1: the case now carries two outputs instead of one — an evidence state and a research priority (level plus stated reason) — and, where a single verdict would mislead by compression, component verdicts for the separable parts of the question. Display governance tightened site-wide: only a human-endorsed assessment run may present as the editorial assessment; unreviewed AI runs are labeled drafts and never silently replace it.

    Why: "Weak evidence, strong reason to investigate" is this site's signature category; a single verdict cannot express it. Priorities and component rows were set editorially in this change and are open to revision on review.

  28. Fable (AI agent) · AI-assisted

    Added the first AI assessment overlay (run orch-or-2026-08-22-fable-1) as an append-only file. Labeled an AI-generated draft; no human has reviewed the claim-by-claim verdicts.

    Why: Assessment overlays are drafts for human editors, never published conclusions. Kept separate from canon so canon stays auditable.

  29. Fable (AI agent) · AI-assisted

    Anchored the Erdős unit-distance evidence on Sawin's independent arXiv paper rather than on OpenAI's announcement. Two direct fetches of the OpenAI page timed out, so the announcement page was never read by this project; its source record says so, and the mathematical fact is carried by an independent working mathematician who states the result and builds on the construction.

    Why: A company's announcement about its own unreleased model is the weakest link available for a load-bearing fact, and in this instance a stronger one existed.

  30. Fable (AI agent) · AI-assisted

    Recorded a documented false-positive history for the AI-mathematics evidence line (ORCH-E027): an October 2025 round of AI Erdős-problem claims collapsed when the solutions were found already in the literature, and a Christmas 2025 claim on Erdős Problem 333 was withdrawn within hours after a 1977 resolution by Erdős himself was identified.

    Why: Symmetry. The microtubule results in this case are annotated with their replication and independence problems, and a newer, more fashionable evidence line gets the same treatment.

  31. Fable (AI agent) · AI-assisted

    Declined to cite the social-media post that prompted the AI-versus-Penrose line. The linked X post returned HTTP 403 to unauthenticated fetch and was never read by this project. The critics' argument is instead carried through verifiable primary sources — Sawin's arXiv paper, the AlphaProof Nature paper, the official IMO grading announcement, and the ICML position paper — with the case editor's own essay cited for the framing of the debate.

    Why: Content policy: an unreadable source is not a source. The underlying evidence line is real and verifiable independently, so nothing was lost by declining.

  32. Fable (AI agent) under Eugene (ejhong) · AI-assisted

    Added the non-computability theme covering Penrose's Gödelian motivation and the recent AI-mathematics results offered against it, at the case editor's request. Scoped explicitly to the motivation rung: ORCH-C040 and ORCH-C041 are assessed as low-diagnosticity for the headline thesis, and both claims state that neither refutation nor vindication would move the microtubule, anesthesia, or collapse claims.

    Why: Critics present AI mathematical performance as refuting Orch OR and proponents treat the Gödel argument as supporting it. Both inferences skip a rung. Making the scoping visible in the ladder is more useful than adjudicating the philosophy.

  33. Fable (AI agent) under Eugene (ejhong) · AI-assisted

    Recorded as a substantive overstatement in the case editor's primer, and corrected in the case's assessment: the primer states that Hameroff's two 2020 predictions "have been confirmed by multiple independent laboratories (Princeton, UCF, NIMS Japan, and Wellesley)." The anesthesia-potency prediction has not been independently confirmed. The two epothilone-B studies share a senior author and research program, and the one genuinely independent stabilizer study (Li et al. 2025, Mass General) found the effect running the opposite direction.

    Why: The editor's own essay is the article's backbone, so its central empirical overstatement had to be handled explicitly rather than inherited. Independence of replication is the specific thing this case tracks most carefully.

  34. Fable (AI agent) · AI-assisted

    CORRECTION to the deep-research report: Warren's comment on the Kerskens result is 2023, Journal of Physics Communications 7(3): 038001, not 2022. The authors' reply (7(3): 038002) was located and verified at the same time.

    Why: The report placed the critique in the same year as the paper it answers. Minor, but the reply's existence matters for representing the dispute as unresolved rather than as a one-sided refutation.

  35. Fable (AI agent) · AI-assisted

    CORRECTION recorded, not adopted as evidence: the primer states that anesthetics shorten delayed luminescence in tubulin while structurally similar non-anesthetics do not, attributing this to Zoghi et al. 2025. The full accessible abstract of that conference paper says nothing about anesthetics. The claim is therefore not carried anywhere in this case, and the discrepancy is recorded on both the source and evidence records.

    Why: This would have been a useful specificity result had it existed — precisely the kind of claim that is tempting to import on trust. If a fuller presentation contains it, it can be added when the document is in hand.

  36. Fable (AI agent) · AI-assisted

    CORRECTION to the case editor's primer: Saxena et al. 2020 was published in Fractal and Fractional 4(2): 11, not "Scientific Reports, 10, 20128". The deep-research report flagged the same error independently.

    Why: Not a cosmetic correction. The claim's weight rests on measurement reliability rather than conceptual novelty, and the misattribution inflated the apparent venue quality of the field's most-cited unreplicated result.

  37. Fable (AI agent) · AI-assisted

    CORRECTION to the case editor's primer: Kalra et al. 2023 is ACS Central Science 9(3): 352–361, not 9(3): 519–529 as the primer states. Volume, issue, and year were correct.

    Why: Page ranges are exactly the field where a plausible-looking error propagates silently. Recorded rather than quietly fixed.

  38. Fable (AI agent) · AI-assisted

    Verified all thirty sources independently before import — DOIs resolved through the Crossref API, abstracts and author lists read through PubMed, publisher pages, the arXiv API, or the official conference program. No citation was imported on the deep-research report's authority alone.

    Why: Deep-research reports fabricate and garble references, and this one supplied most of the case's raw material. Four discrepancies were found, which is roughly the rate that justifies the policy.

  39. Fable (AI agent) under Eugene (ejhong) · AI-assisted

    Case created. Eighteen featured claims across three rungs — five attribution, four mechanism, nine observation — with twenty-nine evidence records, thirty sources, and eight research opportunities, drawn from the case editor's two essays ("Orch OR Ascending", April 2026, and "The Erdős Unit Distance Problem and Penrose Understanding", May 2026) and a commissioned deep-research report on the theory's claims, evidence, replication status, and experimental roadmap.

    Why: Orch OR is a case where the public argument and the evidential situation have come apart. The debate is still conducted as though the question were "can quantum effects survive in a warm wet brain," while the actual evidence has moved to a much narrower and more tractable gap. Mapping the rungs separately is the main contribution.