What looks like noise is often organization we don’t yet know how to read.
I’ve run into some version of that problem repeatedly over thirty years: in human behavior under pressure, in neural signals, and now in markets.
I don’t know whether those are useful analogies or evidence of something deeper. That’s one of the questions driving my work.
The neural question would barely be askable without an instrument. That instrument is the signal-processing platform built by my co-founder, the physicist Prasanta Pal — and much of what this page proposes to test is, in the end, a test of what it can see.
This page separates what is established from what we’re observing, what we suspect, and what we’re building to test it.
I use know narrowly here: established findings with published evidence behind them. My own work appears below and is graded differently. These are degrees of evidentiary commitment, not degrees of importance — something in what we suspect may turn out to matter far more than anything in what we know.
What we know
Established, published, and not by us.
Under sufficient load, human beings can perform below capacities they demonstrably possess under other conditions. Several literatures converge on versions of this pattern: the physiology of allostatic load, neurovisceral accounts of regulatory flexibility, and performance research on choking under pressure.
Regulatory capacities can change with training. Research on contemplative practice and altered traits provides evidence that repeated practice can produce changes that persist beyond the practice session itself.
Living systems hold themselves far from equilibrium. A living system is not at rest and not at a set-point. It maintains an ordered, low-entropy operating configuration against a constant thermodynamic pull toward the inert. The stability living systems achieve is dynamic rather than static.
What we’re observing
Ours. Each item carries its status.
Nervous systems
In preparation
Equilibrium: A Non-Equilibrium Account of the Regulatory Substrate of Human Function
Work conducted through FIELD produced a signal pattern we are investigating as a possible measurable correlate of what we call equilibrium — the per-individual regulatory configuration a person runs from. That interpretation remains a hypothesis. The paper names the construct, grounds it in non-equilibrium thermodynamics and allostasis, and proposes that its structure can be characterized from relational structure retained in the physiological signal, rather than only through scalar summaries.
With coauthors for review prior to submission. Findings have not been peer reviewed. Results and citation will be added here on publication.
FIELD
Dormant
Not currently an active development priority. The research remains part of the program.
Individuals
Ongoing
Thirty years of behavioral practice with people under real load — founders, executives, operators — produced recurring patterns consistent enough to be worth naming and to motivate formal study.
Practitioner observation, not controlled study. It is where the questions came from. It is not where they get answered.
What we suspect
Conjecture. Held openly, and the reason the rest of the work exists.
The larger question is whether some of what we’re seeing at very different scales is related, or merely looks related. That’s one of the things we’re trying to find out.
Similarities across domains are easy to overread. The burden is to show that any apparent relationship survives domain-specific explanation and independent methods of analysis.
We don’t know yet. We think it is worth finding out, and we think the honest position is to say which it is.
What would change our minds
Three things would count against the cross-scale conjecture, and we would rather name them than wait to be shown them.
The same analytic tools — multiscale entropy, detrended fluctuation analysis — can be applied to any time series, and any time series will yield a profile. Similarity of output across domains may be an artifact of similarity of method rather than a finding about the systems. If the resemblance cannot be shown to survive analysis by independent means, it is a property of our instruments and not of the world.
The cross-scale conjecture has no foundation unless the configuration is real at the scale we can measure best. If the structure-preserving signature does not converge across weakly coupled biosignals within one person, the ground floor is gone and nothing above it stands.
If domain-specific explanations account for the observations completely, the stronger conjecture fails. Apparent resemblance across domains is not evidence of a shared underlying principle.
What we’re building to test it
Instruments and experiments capable of telling us we’re wrong.
A richer read of the regulatory substrate. The standard indices used to read regulatory state — heart-rate variability foremost — reduce a rich physiological signal to a handful of summary numbers. Complexity physiology has argued for three decades that meaningful physiological information exists in precisely the structure those summaries discard. Those established tools are being assembled around a different target: a named per-individual regulatory configuration rather than a correlate of aging or disease. The signal-processing work is Prasanta Pal’s; mine is the behavioral model and the construct it’s being pointed at.
The test we could fail. If that configuration is genuinely a property of the whole regulatory system, its signature should converge across biosignals within one person — and specifically across channels that are not physiologically coupled to one another, since coupled channels would agree whether or not a common regulator exists. If the retained structure adds nothing beyond what the summary numbers already share, the proposal is wrong.
Research collaborators
Dr. Prasanta Pal
Doctorate from Yale under Michel Devoret, following undergraduate work at IIT Kharagpur. Postdoctoral research at Yale School of Medicine quantifying blood flow in the human heart from MRI. Formerly Chief of Engineering and Technology at Mindsciences, acquired by Sharecare. Author of dozens of published papers. Inventor on U.S. Patent 12,130,892 B2, “Universal outlier detection and curation of missing or noisy data points,” assigned to Brown University.
He builds the measurement. Intended coauthor on the equilibrium paper.
Attribution, not an advisory board. Papers carry full authorship as submitted.
Where this might lead
This is not a research finding. It’s the horizon that makes the questions worth pursuing.
The period I’m working toward is one I call the Age of Coherence — when we become able to read the health of complex living systems on increasingly common terms, and coherence stops being a metaphor people reach for and becomes a property you can measure.
The nearest-term version of that is not exotic. You cannot build technology that makes people more themselves unless you can tell whether it is doing that, and at the moment we mostly cannot — we measure engagement, not whether a person is becoming more or less available to themselves.
Whether that’s twenty years out or a category error, I don’t know. It’s the reason the work above is worth doing carefully.