NSF Accepted my Grant Pitch for Studying Quantum Brain Effects

The key to understanding how the brain operates so efficiently at only 20 watts of electricity could come from a better understanding of how migratory birds' internal compass helps them navigate seasonal flights.

While attending the University of California, Berkeley in 2010-2012 I was interested deeply in the topic of what distinguishes the human mind from machines, and what endows people with agency. I assumed the physics by which the brain works must necessarily be different than computers to allow free will. I was introduced to the work of Dr. Penrose on the topic, and later concluded under Dr. Vopson's research group in the UK in a publication that to emulate human consciousness would be an NP-hard problem - intractable by both classic and quantum computers - on par within NP-hard post-quantum lattice cryptography. 

Before coming to Colorado, I traveled the country living in a small off-grid trailer I built attached to my car through both the Iowa winter (which was colder than the arctic) and through the Arizona desert. One hot afternoon I went to the science center in Phoenix Arizona where a young woman named Aurora was presenting on the topic of autonomous aerial vehicles and I told her about my situation. She recommended for me to apply to Aurora Flight Sciences - a research and development subsidiary of Boeing where I then got an offer. 

Combining what I learned at Boeing with what I learned in college and with the inspiration from along the way, I conceived of a new technology - the same physics I had thought about at Berkeley is responsible for the internal compass (or "Liahona") of migratory birds which can be appropriated towards developing novel autonomous sensor and navigation systems. I have since published on this topic and presented at the APS conference, and was invited to present at the TSC conference directly by Dr. Hameroff this year. I have discussed this research with Dr. Hore at the University of Oxford and Dr. Thorsten Ritz at the University of California Irvine agreed to partner on this research, and my pitch was accepted by the NSF for a proposal submission. 

The idea is that the information migratory birds sense from their environment that helps them navigate gets encoded onto optical/spin systems in their brain tissues, and this is also the same mechanism that the human brain uses to encode information that performs the equivalent of backpropagation so efficiently across the tissue macroscopically at scale. I have spent the past decade and a half building a research program around that mechanism, both as physics worth finishing and as a sensor worth making that should help autonomous aerial vehicles navigate in GPS denied environments. I want to set down plainly what is solid, what is not, and where I think the genuinely interesting risk sits, because the field tends to blur those three together and the blurring helps no one.

The mechanism, which is the part no longer in serious doubt

A photon is absorbed and drives an electron from one site of a molecule to another. That leaves two unpaired electrons, one on each fragment, and the pair is born in a correlated singlet state, meaning the two spins are anticorrelated in a specific quantum sense. The pair does not sit still. It oscillates coherently between that singlet arrangement and a triplet one, and the clock that drives the oscillation is the difference between the magnetic environments the two electrons feel. Each electron is coupled to nearby nuclear spins through the hyperfine interaction, the couplings differ between the two radicals, and that mismatch is what interconverts singlet and triplet. An external magnetic field shifts the triplet energy levels and so changes how far and how fast that interconversion runs. Singlet and triplet pairs then recombine into different chemical products, so the field leaves its mark as a change in how much of each product the reaction makes.

None of that, on its own, is a compass.

A compass needs direction, and direction comes from the fact that the hyperfine couplings are anisotropic. The rate of singlet to triplet mixing depends on the angle between the molecule and the field, so the product yield carries heading rather than mere field strength. Hold the molecules still and the yield becomes a function of orientation. Let them tumble freely and the anisotropy averages straight back to zero, which is the single most important practical fact in the whole subject.

The biology lands on flavin chemistry. The carrier is cryptochrome, a protein in the retina, and the field acts on a radical pair formed between its flavin cofactor and a chain of tryptophan residues. Henk Maeda and Christiane Timmel and Peter Hore and their colleagues showed back in 2008 that a sub-millitesla field changes the lifetime of a radical pair in a synthetic carotenoid donor and fullerene acceptor triad, the first chemical compass built on a bench. Kerpal and coworkers pushed the directional response down to Earth strength about a decade later. Emily Xu and the Oldenburg and Oxford groups showed in 2021 that cryptochrome four pulled from a migratory robin is magnetically sensitive in a test tube, and more so than the same protein from birds that do not migrate. The mechanism is real, it is specific, and it has names attached to every step.

What is not solved, and why that is the opportunity

Here is the gap that I think is worth a career. Every clean directional measurement to date has required freezing the sample, somewhere down near a hundred kelvin, precisely to stop the tumbling that would otherwise erase the anisotropy. The bird does this at body temperature inside a structured protein, and we cannot yet reproduce that on a bench without the cold. The field effect that survives at room temperature in the lab is a percent level change in fluorescence, beautiful and real but a long way from a heading you would trust over open water.

Many so-called "quantum theories of mind" like those described by Dr. Penrose often paper over this gap, famously criticized by Max Tegmark as implausible in brain tissues - as the brain is warm, wet, and noisy which is inhospitable for macroscopic quantumlike effects to occur. Understanding how the avian compass works provides an avenue for probing this mechanism at play where controlled information about the environment is fed through the tissue.

So the near term target of the program is narrow and concrete. Build a calibrated, reproducible, room temperature optical readout of the radical pair field effect in cryptochrome and flavoprotein systems, with no cryogenics and no magnetic shielding beyond ordinary background control, and state its field resolution in real units rather than in arbitrary contrast. The field effect surfaces in fluorescence yield and in fluorescence lifetime, and lifetime is the better handle because it is far less sensitive to how much sample you have and how bright your source is. Time resolved methods, and phasor analysis in particular, can pull a small lifetime shift out of a messy signal without committing to a fitted model in advance. Akira Antill and Maeda reached roughly two parts in a thousand on fluorescence near the single photon level in 2025, which tells you the sensitivity is within reach and that nobody has yet turned it into an actual fielded transfer function at room temperature. That transfer function, yield and lifetime against field magnitude and orientation, is the deliverable. A clean negative would be informative too, because it would set an honest bound on how far the optical route can go.

This is not a confirmation of someone else's result. A room temperature, field referred lifetime transfer function for a flavoprotein radical pair, with resolution stated in tesla per root hertz, does not exist in the literature. It is the first measurement of its kind, it is publishable on its own terms, and it is the rung that everything more ambitious has to stand on.

The frontier, fenced off honestly

Now the part I find hard to stop thinking about, which I will mark clearly as hypothesis so that no one mistakes enthusiasm for evidence.

The readout above treats the emitted light as a meter needle, a brightness or a decay rate that we read off. The deeper question is whether the light carries more than that. The spin state of a radical pair has structure beyond a single yield number, structure that in principle could appear in the polarization of the photons it emits, and a growing body of theoretical work asks whether dense networks of biological emitters can radiate cooperatively rather than independently. If field dependent spin information is genuinely imprinted on emitted or collectively emitted light, then light stops being a passive meter and becomes a channel, something that moves information from one place to another inside tissue.

That idea has a clean and falsifiable first test, and I value it precisely because it can fail. If the magnetoreception signal is encoded on emitted light, then selectively blocking or scrambling that light should measurably degrade the field response, while leaving the rest of the chemistry intact. An intervention and a readout, with a real null available. That is an experiment, not a manifesto.

I will say the harder thing out loud, because the field needs people to say it. Showing that light carries a field dependent signal is not the same as showing that anything computes with that signal.

A wire carrying a voltage is not yet a circuit doing arithmetic. There is a long and genuinely speculative arc that runs from optical signaling, to the idea that such signaling underlies the strange efficiency of neural tissue, to the much larger claim that something in the brain implements the equivalent of error driven learning by optical means. I find that arc compelling as motivation. I do not present it as established, and neither should anyone else, because backpropagation in the brain is unresolved by any mechanism at all, electrical or optical, and you cannot demonstrate that light implements a process that has not been shown to occur in the first place. The honest structure is to treat the encoding experiment as a necessary precondition for the larger story and nothing more. If light carries no functional signal, the larger story is dead. If it does, you have earned exactly one step, and the next step is a separate and much harder experiment.

Why this is important

Because the payoff at the near end is real and the payoff at the far end is enormous, and the two are connected by a sequence of measurements you can actually run rather than by assertion. A passive, room temperature, unshielded magnetometer with no moving parts would matter on its own, in navigation that cannot lean on satellites and in sensing where cryogenics and heavy shielding are not options. And the same readout, made sensitive enough, is the instrument you would need to ask the wilder question at all.

The bird already solved this. It runs a quantum mechanical measurement in warm, wet, disordered tissue, reliably enough to bet its life on twice a year amplified to a macroscopic environment. We have not matched that on a bench, and we do not fully understand how it is done. Closing that gap is worth doing carefully, one rung at a time, with the speculative weight kept off the rungs that have to bear load. That is what this program is, and that is the order I intend to build it in.