Alongside the products, the lab runs a small research practice. The method is always the same: take a real scientific question, build it into a map of the published literature, find the cheapest experiment that could actually settle it, run it, then have independent reviewers try to tear the result apart before anything is called a finding. The rule is the same as everywhere else here — separate fact from interpretation, cite the real sources, say “contested” when it is contested, and report a null result as plainly as a positive one.
A from-scratch investigation of the “quantum consciousness” idea: a literature map, then cheap discriminating tests at the three weakest points of the leading theory. Honest verdict — not supported on current evidence, but not disproven.
A re-audit of celebrated empirical “laws” — the neat scaling relationships from the textbooks — against the full set of open datasets they claim to govern, with proper model comparison rather than a fitted line on a log–log plot. A clean confirmation and an honest “it doesn’t hold as widely as we thought” are equally worth reporting.
A careful, cited read of where the “quantum consciousness” question actually stands — built as a literature map plus a set of cheap, discriminating tests, each one independently fact-checked twice.
The popular claim is that the brain is, at some level, a quantum computer — that quantum effects in neurons do real work behind thought and consciousness. The best-developed version of that idea is the Orch-OR program (Penrose and Hameroff): microtubules inside neurons run quantum computations that are terminated by a proposed gravitational form of wave-function collapse.
This study took that program apart at its three load-bearing points and asked, at each one, the cheapest question that could actually discriminate quantum from classical. The honest result is that the Orch-OR-style quantum-brain program is not supported on current evidence — its foundation, its main empirical claim, and its keystone “direct evidence” each fail, independently, to clear the bar. A second, better-built quantum-brain hypothesis (Fisher’s nuclear-spin proposal) survives the decoherence problem but is so far unconfirmed, with a clean isotope test already returning a null.
Three things are worth saying plainly, because they are easy to blur:
This page is the plain-language summary. For the full treatment — the decoherence mathematics shown and derived, complete methods, the per-probe results with all the numbers, an honest limitations section, and the full numbered references with venues and DOIs — read the comprehensive technical report (also available as a print / PDF version).
The program rests on three supports, so the study built one discriminating probe for each — its foundation (can a quantum state even last long enough?), its main empirical handle (does the anaesthesia evidence actually need quantum mechanics?), and its keystone (does the one piece of “direct evidence” really show entanglement?). Each was attacked at its strongest point. These are compute-and-literature probes, appropriate to an independent lab with no wet bench — not new lab experiments.
A conscious “moment” (40 Hz gamma) needs a state to last on the order of 25–100 ms. This probe re-implemented both published microtubule decoherence formulae — Tegmark’s and Hagan’s — from the papers’ own equations, reproduced each paper’s own anchor, and plotted both on one axis against that threshold.
Both rival numbers fall short. Tegmark’s ~10−13 s is about twelve orders of magnitude too brief. Hagan’s most generous published ~10−4 s is still roughly 250× short of 25 ms. The entire dispute hinges on one parameter — the superposition separation. The gap can just be touched only by inserting Orch-OR’s own nuclear-scale (~1 Fermi) separation into the decoherence formula, which is exactly the move physicists most dispute — and even then it still misses 100 ms.
Verdict Unbridged. The foundational requirement is not met on the published numbers; the gap closes only on the single most-contested assumption.
Anaesthetics reversibly switch consciousness off, so whatever they touch is plausibly part of the substrate — and proponents point at microtubules. This probe pulled 13 anaesthetic–microtubule studies into one effect-size read, sorting each into “classical-compatible” vs “requires a quantum observable.”
Strip the word “quantum” out and the data lose nothing. The strongest result — an anaesthetic-delaying microtubule effect at Cohen’s d = 1.9 (Khan 2024) — is a clean behavioural result that says nothing about any quantum mechanism. Zero of six microtubule–anaesthesia results require a quantum observable. The one genuinely quantum candidate — a xenon nuclear-spin isotope effect — is real and sizeable but unreplicated in eight years, mass-confounded, and points away from microtubules entirely.
Verdict Confirmed (high). A classical microtubule role explains every result; “quantum” is interpretation, not data.
The piece most often cited as direct evidence is a 2022 MRI “entanglement-witness” signal — heartbeat-locked, present only when awake (Kerskens & Perez). This probe reconstructed the witness logic and built a classical null-model at the paper’s own sequence parameters.
The witness theorem is valid, but it only bites if the signal is demonstrated to be genuine entanglement — and ordinary long-range intermolecular NMR coherence between water molecules produces exactly that class of signal with no entanglement at all. A discriminating point neither camp had drawn: the signal vanishing at the “magic angle” (54.7°) confirms a dipolar origin but does not separate classical from quantum — both carry the same angular factor and both vanish there.
Verdict Does not discriminate. Contested, single-lab, unreplicated — not debunked, but not confirmed.
After Orch-OR, the study probed the other serious candidate: Matthew Fisher’s proposal that phosphorus-31 nuclear spins in “Posner molecules” carry quantum information. This is the best-formulated and most genuinely testable quantum-brain hypothesis — and crucially, a nuclear-spin substrate is a real answer to the decoherence problem that sinks Orch-OR, because nuclear spins are intrinsically far better isolated.
But the evidence leans unsupported. The one clean discriminating isotope test run so far (calcium-43 vs calcium-40, 2020) came back null against the prediction; the supporting lithium-isotope results carry the same mass confound that weakens the xenon result; and the molecule’s carrier structure is itself contested.
Verdict The better-built, falsifiable hypothesis — genuinely alive, but unconfirmed, with a clean null already on the board.
| Pressure-point | Probe | Verdict |
|---|---|---|
| Foundation can a quantum state last ~25 ms? |
G1 | Unbridged. Both decoherence numbers fall below threshold; the gap closes only on the most-contested assumption. |
| Main empirical handle does anaesthesia need quantum? |
G2 | No (high confidence). A classical microtubule role explains every result. |
| Keystone “direct evidence” does the MRI signal show entanglement? |
G3 | Does not discriminate. The signal is what ordinary intermolecular NMR coherence produces with no entanglement. |
| Other hypothesis Fisher nuclear-spin / Posner |
G6 | Best-built and testable — but unconfirmed; a clean isotope test already returned a null. |
This is the essential calibration. Genuine, measurable functional quantum effects in living systems do exist — they are just narrow, and not about the brain:
Two lessons follow. Real quantum biology operates at femtosecond–picosecond scales — roughly ten to thirteen orders of magnitude shorter than the millisecond cognition Orch-OR needs, which is exactly the gap the foundation probe quantifies. And the discipline’s own flagship was over-interpreted and corrected, so any microtubule-coherence claim inherits that cautionary precedent: a real spectroscopic signal is not the same thing as functional coherence.
Kept honestly separate from what looks closed: the Kerskens brain-entanglement signal is open pending independent multi-centre replication and a positive classical forward model (or a separability bound the signal provably exceeds); the xenon nuclear-spin anaesthetic effect is open pending a mass-corrected isotope replication; and Fisher’s proposal remains a specific, partly-testable, but unverified island. The deepest open question is whether any of these yields a prediction that discriminates from a classical alternative at all — which is precisely what the next steps target.
On current evidence there is no good reason to think quantum mechanics does functional work in the brain. The leading theory (Orch-OR) is not supported — its foundation, its main empirical handle, and its keystone evidence each fail. The most testable rival (Fisher) is the one to watch, but it is unconfirmed and already carries a clean null. Real quantum biology is established, but it is narrow, ultra-fast, and not about consciousness. Throughout: “not supported” is not “disproven.”
Load-bearing references only. The full study cites 35 papers across a 47-node knowledge graph; every load-bearing number was independently checked on the live literature, and each probe was fact-checked by two skeptical reviewers.
The literature was mapped into a knowledge graph, each probe was run as transparent compute or close reading, and every load-bearing claim was independently fact-checked twice. Retrieved paper text was treated as data, not instruction. No numbers, citations, or results were invented; figures that could not be re-derived are attributed to their source paper rather than asserted. Evidence grades are deliberately conservative — a real signal with a contested interpretation is graded contested, never established. The work was done with the help of an AI research collaborator, with the judgement and the standard kept by the lab.