That perspective gives him something rare. A complete picture of how complex this problem really is—and a clear-eyed appreciation of where the real leverage points might be.

Tom sees mental health as a systems design problem. The complexity of the system has to match the complexity of a single patient’s needs. And nothing moves without solving the economics. As he puts it, until you change payment, you don’t change adoption.

It’s a fascinating conversation with someone who sees the full weight of the challenge—and believes mental health treatment may be at a tipping point.

•

This podcast is for general educational and informational purposes only. Sensorium’s compounds are investigational and have not been approved by the FDA for any use; nothing discussed here should be taken as a claim about their safety or efficacy. The views of any guests are their own and do not represent Sensorium. Nothing in this podcast is medical, legal, or investment advice, and nothing here is an offer to sell or a solicitation of an offer to buy any securities. Statements about future plans, timelines, or outcomes are forward-looking, reflect current expectations only, and are subject to change.

Subscribe now

Put yourself back in early 2021 for a second. We were getting vaccinated, going back to in-person meetings, and standing in rooms where nobody was sure whether it was rude to take the mask off or rude to leave it on. Feels strange to go back to that time but something else was happening that you might not have been aware of.

Almost drowned out by the COVID narrative, there was a lot of enthusiasm emerging for psychedelics. Companies were entering the clinic touting the promise of help with intractable challenges — labels for patients with diagnosis of PTSD, treatment-resistant depression, severe anxiety disorders. Private capital was interested for the first time in a couple of generations or more. Public capital followed that with the IPO of some of the first psychedelic-focused companies.

Look at where those companies are now. A good number of them have been acquired and have compounds that may be nearing approval. (See for example, my substack on Atai Beckley). Whatever you thought of the state of psychedelic science in 2021 (there were and still are a lot of inspiring mysteries), and whether the investments were warranted, the clinical data have been impressive and the market has voted with real money since.

The long way around

Drug development in psychiatry is hard. (Let me rephrase—all drug development is hard; psychiatry has unique elements that can make it even harder.) What is compelling about psychedelics is their rich history. (See my post on the studies of LSD and its first go as a pharmaceutical drug.) When humans already know how a substance acts on the brain in terms of the subjective experience, we have a leg up on how to turn it into a medicine. This flips modern discovery on its head and is more aligned with earlier empirical versions of drug development. These days we often start from a biological discovery, build an animal model of disease to test, and eventually work our way back toward a person. In psychiatry this means we can spend a decade finding out whether the model was right. Sensorium’s vision, which isn’t novel per se (see how even new drugs for schizophrenia like Cobenfy come from this line of thinking), was to start from the person and reverse engineer to mechanism.

So back in 2021, my co-founders and I were surprised how much adjacent space to psychedelics was being underserved. What do I mean by adjacent? Psychoactive but not hallucinogenic molecules from plants. We were drawn to some amazing plants with a key feature: real psychoactivity and centuries of human use behind them, where you had a sense of what might be possible if you could only figure out how they might be working, so that you could improve their likelihood of success in a modern medicine context.

There are twenty thousand or so psychoactive plants… how do you pick?

The early challenge we needed to solve in 2021 was not “how do you find interesting opportunities?” It was, “how do you know which opportunities are most likely to succeed?”I won’t go into the details here, but we were fortunate to launch during the early days of the LLM technologies that are now ubiquitous. We built a highly curated, bespoke knowledge graph (one that now has more than 1.5 billion encoded data relationships). Seems like overkill, right? Not if you’re looking for patterns: patterns of knowledge rediscovery, connections between data sets that are hidden unless you know how to “wire” the graph together.

Our first clinical readout

Last week, Sensorium’s chief medical officer was at Psych Congress, sharing clinical data from one of our programs for the first time in the company’s history.

I won’t walk through the data here. The press release is linked here if you’re curious. What I will say is that it suggests we may be onto something with this concept. Across all of our programs we’ve seen that we can discover unique biochemical mechanisms, we can increase the hit rate in translational models, we can maintain the magic of the natural compound while we improve it. These early clinical data are the first signals that bring us full circle. We started from human observations and are back to human observations. That connectivity — starting from knowledge that dates back hundreds or even thousands of years and connecting it to the most modern neuroscience tools — feels very special to me.

This is a special moment in psychiatry

I am so excited to live at a time when there will likely be a step function change in our understanding and treatment of psychiatric illness. Over decades, medicine has built a knowledge base and tools across all of CNS, and that investment is really starting to pay off, particularly for psychiatry. My news feed is filled with the amazing progress and trial success that has been absent for nearly all of my life. I’m grateful to be part of this moment, with a team that cares so much about helping people who live every day with a mental health disorder.

I have ended many posts by saying, “I choose optimism.” Perhaps the overwhelming feeling I have today is, “I no longer have to choose.” The signals are all around us.

Leave a comment

What happens when the diagnostic categories anchoring decades of psychiatric research were never designed to describe anything real? Steve calls this psychiatry’s epistemic self-injury. The DSM’s criteria for depression, schizophrenia, and bipolar disorder were built to maximize agreement between clinicians, not to pick out something true in nature, and their architects never claimed otherwise. Thirty-five years of case-control MRI studies built on those labels have produced nothing clinically useful. And when Steve compares the genetics of schizophrenia and bipolar disorder, the two correlate at close to seventy percent, evidence that what we treat as separate diseases may be one shared biology wearing different names.

So what’s the fix, if not abandoning genetics altogether? Steve’s answer is to work backward: cluster patients by genetic and symptom correlation first, then rebuild diagnostic categories from the ground up. At the Broad, that means running a study almost no one else would attempt, tracking psychiatric patients for five years with repeated lumbar punctures to sample cerebrospinal fluid, guided by a single genetic clue. A complement gene tied to schizophrenia risk turned out to flag weak synapses for elimination, implicating the same pruning process that reshapes the brain during adolescence. It is a hypothesis borrowed from Alzheimer’s research, chasing a biological anchor psychiatry has never had.

We close with Steve’s case for what it will take to break the field’s stuckness: bringing industry and regulators into the same room as researchers, and his hope that real biomarkers, not another diagnostic manual, are what finally move psychiatric treatment forward.

•

This podcast is for general educational and informational purposes only. Sensorium’s compounds are investigational and have not been approved by the FDA for any use; nothing discussed here should be taken as a claim about their safety or efficacy. The views of any guests are their own and do not represent Sensorium. Nothing in this podcast is medical, legal, or investment advice, and nothing here is an offer to sell or a solicitation of an offer to buy any securities. Statements about future plans, timelines, or outcomes are forward-looking, reflect current expectations only, and are subject to change.

The paper is based on a very rare gift.

Human brain data from implanted electrodes in the hippocampus and the visual cortex. In my brain imaging work, I often described the value of non-invasive imaging techniques given that not too many people volunteer to let me sample or put a probe in their brain. This is a rare exception because the probes were clinically necessary.

The patients that participated have drug-resistant focal epilepsy and were undergoing pre-surgical evaluation to decide whether resective surgery will help — and where to cut. Neurosurgeons implanted stereo-EEG electrodes to record seizures directly from inside the brain and localize the seizure focus.

While the patients were awake, watching streams of images — some predictable, some unexpected — the hippocampus was doing something no one had directly caught inside a living human brain before.

Right before a surprising image appeared, sharp wave ripples in the hippocampus fired more often and lasted longer than they had before the predictable ones. And those ripples reached ahead in time, into the visual cortex — quieting the background static, priming the brain to respond more sharply to whatever was about to arrive.

The memory system, in other words, was pre-loading perception. Before the surprise. On the order of milliseconds. It isn’t guessing what will come next. It’s turning up sensitivity perhaps like you lean toward a speaker in a crowded room so you hear what they’re about to say. I find this beautiful!

When I think of the hippocampus, I think about memory (aka the past). What happened. Where I was. Who I met.

So is the brain seeing the future?

The short answer is no. The brain isn’t seeing the future. The brain just noticed, from the last few seconds, that it’s in a stretch where surprises keep happening and it’s bracing for the next one, without knowing what that surprise will be. Think of watching a horror movie (my favorite I think is still The Ring). You don’t know exactly when the scary thing is going to jump out, but you know something is coming and your brain is trying to be ready!

Subscribe now

The implication here is super important.

Prediction, uncertainty, and memory are not three different systems. They are all part of one.

On reflection (pun from the water image intended), forward prediction is exactly what memory is for. There is no evolutionary reason that we need to store the past unless the stored material is going to help with the future.

We are all bracing for what’s next in a world increasingly full of noise…

What happens when you ask thousands of people about trauma instead of trying to fit them into a diagnostic box? Kerry found that nobody fit the DSM—everybody had everything. Once he started asking specifically about trauma, he discovered it was everywhere, hiding beneath substance use, depression, and anxiety. He began thinking about trauma the way physicians in the 1800s thought about syphilis—a hidden driver behind conditions we were attributing to something else entirely.

That observation led him deep into the neuroscience of fear memory: how it forms, whether it can truly be extinguished or only suppressed, and whether the hours and days immediately after a traumatic event represent a critical window—like the window we exploit after a heart attack or stroke—to prevent chronic PTSD before it takes hold. And it raises a question that is as much ethical as scientific: if we could dampen the emotional weight of a traumatic memory, should we?

We close with Kerry’s hard-won optimism that as neuroscience and psychiatry converge, more personalized and effective treatments for fear and trauma are finally within reach.

•

This podcast is for general educational and informational purposes only. Sensorium’s compounds are investigational and have not been approved by the FDA for any use; nothing discussed here should be taken as a claim about their safety or efficacy. The views of any guests are their own and do not represent Sensorium. Nothing in this podcast is medical, legal, or investment advice, and nothing here is an offer to sell or a solicitation of an offer to buy any securities. Statements about future plans, timelines, or outcomes are forward-looking, reflect current expectations only, and are subject to change.

$12,000,000,000,000 (yes, that’s $12 trillion in 2024 US dollars).

When suicide deaths attributable to depression are included, it climbs to $14 trillion. That is 0.46% of global GDP.

Every year. For 25 years.

The United States alone accounts for $4.1 trillion. China, $2.1 trillion. The UK, $475 billion.

THINK ABOUT THAT. Depression will cost the world an amount equal to roughly the size of Germany’s entire economy, compounding, for the next quarter century. And the driver is not healthcare spending. It is lost labor force participation. Depression removes people from work. That is the main cost driver.

This new publication posted three months after I wrote about the prevalence data: 1.17 billion of us living with a mental disorder, mental illness now the leading cause of years lived with disability on Earth. That was the mental illness population ledger… this new paper is the price tag for depression (which by the way is actually #2 on the world mental disorder list).

This new publication comes one week after Eli Lilly announced a $2.8 billion acquisition of Atai Beckley, principally for one asset best I can tell. BPL-003, an intranasal synthetic 5-MeO-DMT for treatment-resistant depression, with FDA Breakthrough Therapy Designation and Phase 3 readouts due 2029. Bold.

If you are not familiar, 5-MeO-DMT is arguably the most potent psychedelic ever taken seriously in a clinical setting — maybe even the most potent period (someone will probably correct me). Faster, harder onset than psilocybin. Shorter (roughly 15–30 minutes of active experience, versus 4–6 hours). Dramatically more intense. Users describe a near-complete dissolution of the self. The boundary between “me” and “everything” just gives out for the length of the trip. Some report that the toad (another name for the molecule given its origins) is the closest thing chemistry has produced to a mystical or religious experience. Some report that it is the closest thing chemistry has produced to being terrified out of existence. Either, apparently, could be key to the therapeutic antidepressant effect. The Phase 2 data show rapid and durable antidepressant benefit after a single administration. But man, how is this going to work?! (If anyone can figure it out perhaps Lilly can — they certainly have the resources right now and I do hope this works.)

We will have to hold two things at once again — that the intensity of the experience is both the point and the risk. The next few years will be wild.

Subscribe now

Her work traces psychiatry’s recurring pivots. The field abandons one framework for another, convinced this time it has found the real source of mental suffering. What makes Anne’s perspective invaluable is her refusal to let any single story win. Mental illness doesn’t map neatly onto what we call disease. Yet we’ve narrowed our response almost exclusively to medical intervention.

The question she keeps returning to cuts deeper than methodology. What matters most—getting the label right, or addressing the suffering? Should we wait to name something until we see what intervention actually helps? And why have we decided that only one domain of expertise should be involved when the human brain is profoundly social and hardwired to consume symbols and meaning?

This is a conversation about what we’re really trying to solve—and whether science alone can get us there.

•

This podcast is for general educational and informational purposes only. Sensorium’s compounds are investigational and have not been approved by the FDA for any use; nothing discussed here should be taken as a claim about their safety or efficacy. The views of any guests are their own and do not represent Sensorium. Nothing in this podcast is medical, legal, or investment advice, and nothing here is an offer to sell or a solicitation of an offer to buy any securities. Statements about future plans, timelines, or outcomes are forward-looking, reflect current expectations only, and are subject to change.

DT120 is lysergide. LSD. Yes, that LSD.

Subscribe now

Maybe you know that already. But I bet you didn’t know…

From 1947 to 1965, LSD was a legal prescription medicine. The Swiss pharma giant Sandoz sold it under the brand name Delysid, in 25-microgram tablets and 100-microgram ampoules for injection. The official package insert listed an indication straight out of a modern outpatient psychiatry clinic: “analytical psychotherapy” — specifically, the insert read, for use “in anxiety states and obsessional neuroses.” The recommended dosing started at 25 µg and titrated up by 25 µg each session until the patient hit an optimal dose, usually between 50 and 200 µg.

Sandoz mailed Delysid free to any doctor who asked for it for research. Apparently thousands asked because by 1965, roughly 1,000 clinical papers had been published — for depression, anxiety, OCD, alcohol use disorder, end-of-life dread, chronic pain.

Psychiatry was moving.

Then, in 1968, the DEA made LSD Schedule I. Not because the clinical data was bad. Because the culture was anxious. The trials stopped.

With the Definium data I feel this paradoxical sense of both optimism and sadness. Why sadness? Because we were already on this track. We did not lose this molecule to a failed trial. We lost it to a moral panic or political/cultural control that had nothing to do with the clinic or the science. And this is a pattern — hopefully one that we’re about to break. But I have my concerns given that we are not, as a field, very good at holding two things at once — that a treatment can be useful and dangerous. We pick a side. A decade later, we pick the other side. Lobotomies. Insulin shock. MAOIs. Benzos. SSRIs. Now LSD, again. Context and appropriate use matters. Nothing is black in white in psychopharmacology.

What else have we thrown out or overlooked?

As I’ve said in the past, I choose optimism and the hope that we will continue to rediscover and drive old concepts forward with the beauty and depth of modern science. We have a lot to learn from the past — if we listen.

Bryan’s path to neuroscience began with his mother’s schizophrenia and his sister’s experience with LSD—experiences that set a teenager on a lifelong quest to understand how molecules change minds. Decades later, he finally saw what he’d been chasing since he was 14: the crystal structure of LSD bound to a serotonin receptor. “We had a lab meeting, and they showed it, and I started weeping. This is what I’ve been chasing my entire life.”

What makes Bryan remarkable isn’t just the breakthroughs—it’s his ability to hold paradoxes lightly where others see rigid boundaries. He moves seamlessly from Zen meditation to receptor pharmacology, from the molecular to the existential. And his take on mental health research is characteristically provocative: “We’ve been doing an uncontrolled chemical experiment on the entire human population for millennia.”

This is a conversation about pure scientific passion—and why understanding the brain might require us to quiet our own minds first.

•

This podcast is for general educational and informational purposes only. Sensorium’s compounds are investigational and have not been approved by the FDA for any use; nothing discussed here should be taken as a claim about their safety or efficacy. The views of any guests are their own and do not represent Sensorium. Nothing in this podcast is medical, legal, or investment advice, and nothing here is an offer to sell or a solicitation of an offer to buy any securities. Statements about future plans, timelines, or outcomes are forward-looking, reflect current expectations only, and are subject to change.

May is a month when I think about the people I love who have lived through major mental health issues.

I’m guessing mental health is personal to you, too. To almost every single person reading this. But we don’t talk about it even though stigma has gotten better. We have created a quiet us — folks impacted by mental health issues.

You want to know just how many are in the quiet us?

This week, The Lancet put some number behind the quiet us. The Global Burden of Disease 2023 report — 204 countries, using three decades of data, and performed by a massive research team (nearly 1200 people listed as contributors to this paper!). The team found that 1.17 billion people are living with a mental health disorder. That number has nearly doubled since 1990. “Well, there are just more people…" Nope. This was after adjusting for population growth. The rate is up 24%.

In 1990, mental illness was the 12th-leading cause of disability-adjusted life-years globally. In 2023 it was the 5th. It is now the single leading cause of years lived with disability on Earth.

Some context and texure on Mental Illness from the Lancet

Within mental illness, the leading symptom by a meaningful margin, in nearly every region of the world is now ANXIETY. Anxiety disorders are the single largest contributor to global mental-health burden, with major depression second and schizophrenia third. The Lancet team broke this down by region. Major depression only edges ahead of anxiety in East and South Asia and a couple parts of Eastern Europe.

The peak burden lands in the 15-to-19 age group. Women carry more than men, in every region, at every age. It makes me so sad to think about each individual suffering and also the aggregate pain of what these maps represent.

A companion paper from the same research group found that in 90 countries, fewer than 5% of people with major depression receive what the WHO calls “minimally adequate” treatment. Not great treatment. Minimal. We have known this for years and it hasn’t changed.

So what exactly have we been doing every May? Does this mental health awareness month even matter?

We have run, with somewhat extraordinary efficiency, a generation-long campaign to make sure people understand they are suffering. At least in the US, we have made progress on the stigma of mental health. The progress on awareness is important.

But awareness is not treatment.

Awareness is a key peice of the diagnostic step — flawed as that might be (see my essay on the DSM). The numbers in The Lancet are what happens when a society gets very good at the diagnostic step and very bad at everything that should come next.

We are in the middle of a soft revolt against “chemicals.” You can hear it in wellness podcasts and supplement ads, in TikTok monologues about “natural” healing, in the way bright, well-meaning people lower their voice a half-step when they mention being on an SSRI, as if confessing in a Catholic church. In my last Aether in Mind conversation, Jerry Rosenbaum — one of the early psychopharmacologists, someone who helped build the modern field from the inside — told me he was looked down on for that work even by colleagues, early on. Decades later, the stigma has shape-shifted. It now dresses itself up in wellness language and “do your own research” language. This MAHA madness is so frustating (even if it’s well-intentioned, it is so short-sighted.)

Not everyone in distress needs a chemical. Some need a therapist, a community, a different job, a different relationship, a different country to live in.

But…

In many, many cases and certainly more than the current cultural mood is willing to admit— these drugs save lives. They alter trajectories. I have watched this happen up close, more than once, in my own family and among friends. I have also watched it fail — watched the side effects that suck the life out of person, watched someone cycle through five medications before finding something that they could tolerate and that helped. The drugs are not perfect. They need a generation, maybe more, of innovation. I am working on some of that. So are a lot of other people. It cannot come fast enough.

But the chemicals are not the enemy! They are not the lazy choice or the corporate choice or the unspiritual choice. They are a tool — one tool, in a tool kit that has to also include CBT, community, economic security, and a society less designed to break us. We need all of it.

Subscribe now

There’s an possibility this Lancet data reprents the rise in awareness. That is what awareness working looks like. But some of it is real change in the world and the human relationship to it — pandemic, screen time all day, isolation, climate dread, political uncertainty, wars everywhere… take your pick — and we cannot cleanly separate the two. What we can tell is that the suffering is bigger than the current mental health infrastructure, and the gap is widening.

I honestly don’t know what to do with this. On this last weekend of the mental health awareness month, I’d offer, in place of or perhaps in addition to a green ribbon:

• Call one specific person you’ve been worried about. Not “thinking of you.” Just call and check on someone. Today.

• Talk about medication. If you take a psychiatric medication and it is working, consider telling someone — they might need to hear it. If you don’t take one, ask someone who does what it’s been like, and just listen. The stigma dies one conversation at a time. (Totally talk about what drug(s) haven’t worked too! Just normalize the discussion.)

• Donate to a group actually doing the work. Crisis Text Line, The Jed Foundation, and The Trevor Project all hold 4-of-4 stars on Charity Navigator. About $25 funds one trained-counselor conversation with a young person in crisis.

What would it take, in your life or your community, to spend May 2027 not raising awareness — but closing a gap?

Aether in Mind is a newsletter about minds, medicine, and the spaces in between. If this hit something, share it with one person who needs to read it. That’s how we move.

Share

Jerry’s journey began in the 1970s as what he calls a “rogue trainee”—one of the first to embrace psychopharmacology when giving patients “chemicals” was considered radical. From those controversial beginnings, he helped build what became a network of 50-60 specialized clinical and research programs at MGH.

His framework for understanding our current moment is both elegant and unsettling: anxiety, he says, is spreading like an autoimmune disease—a protective system now turned against us, keeping us in a state of arousal that limits our possibilities as humans.

This is a conversation about psychiatry’s hard-won wisdom and the molecules nature left for us to discover.

•

This podcast is for general educational and informational purposes only. Sensorium’s compounds are investigational and have not been approved by the FDA for any use; nothing discussed here should be taken as a claim about their safety or efficacy. The views of any guests are their own and do not represent Sensorium. Nothing in this podcast is medical, legal, or investment advice, and nothing here is an offer to sell or a solicitation of an offer to buy any securities. Statements about future plans, timelines, or outcomes are forward-looking, reflect current expectations only, and are subject to change.

We also got into why most brain medicines still come from plants and fungi.

Here is a number that did not make the recording. Goodwin et al. (J Psychiatr Res, 2020), using U.S. national survey data from 2008 through 2018, found that past-year anxiety nearly doubled among 18- to 25-year-olds in a decade, from 7.97% to 14.66%. Across all adults the rate climbed roughly 30% over the same period. This is pre-pandemic. I often get the question about whether anxiety rates will go back down the farther we get away from the initial wave of COVID-19. I do not see that happening.

Subscribe now

If the brain is a prediction machine, and prediction machines are tuned to the stability of the world they run in, then a decade in which the texture of daily life (work, news, social ground truth, attention itself) got measurably less stable should produce something close to that signal. It did. Anxiety is the system doing its job. The job is just not pleasant when the world refuses to settle.

(I do want to be careful. Trends in self-reported anxiety include real biology, real environment, and changes in screening, vocabulary, and stigma. The effect size in young adults is too large for screening shifts alone, but reasonable people will weight those terms differently. At least that’s my read.)

Listen to the conversation if you want the longer version of the thesis, and the chemistry stories that go with it.

Subscribe now

Every neuroscientist and philosopher who has wrestled with the hard problem of consciousness (i.e. why does subjective experience exist at all?) is working from inside the bottle. Pollan’s book was fantastic at opening my aperture to the theories of consciousness and also to the possibility that consciousness may be the one scientific problem where being human is more liability than asset.

The timing of a paper published last month in Nature Neuroscience was a fortunate coincidence and has me thinking about how AI might be able to get around some of the challenges. Since I do not (yet) believe that AI is conscious, perhaps we can get a look at ourselves, our label, from "outside the bottle”.

What they did

Daniel Toker and colleagues at UCLA built an adversarial AI framework — two models pitted against each other in a kind of game. One learned to detect levels of consciousness from EEG and electrophysiological brain recordings. The other generated biologically realistic simulations of conscious and unconscious brains. The two were then set against each other: one generated simulated unconscious brain activity, the other tried to distinguish those simulations from real conscious brains. Each round, both got better — and what emerged from that competition, without either model being told what to look for, were testable predictions about what actually goes wrong in coma, vegetative states, and minimally conscious states. I think this kind of hypothesis generation is a great use of AI in science right now.

The training set was interesting: over 680,000 ten-second neuroelectrophysiology samples from humans, monkeys, rats, and bats. The AI wasn’t given a theory of consciousness to test. It just learned the patterns — from the outside.

Without being programmed to, the model identified two specific circuit disruptions it predicted would drive unconsciousness: selective damage to the basal ganglia indirect pathway, and abnormal inhibitory wiring in the cortex. Both were then validated: a basal ganglia finding in diffusion MRI from 51 patients with disorders of consciousness, and cortical wiring changes in RNA sequencing of resected brain tissue from 6 human patients with coma. The model also flagged subthalamic nucleus stimulation as a potential therapy — and provided preliminary EEG support from 6 patients with cervical dystonia, though these were conscious patients, not DOC patients. The clinical leap remains large.

Before you think we’re about to solve the hard problem of consciousness…. important caveats

I find this work really interesting, but a few things give me pause — not so much as critiques but more so about where the science is currently. For example, the MRI validation cohort is 51 patients which is meaningful, but small for claims of this magnitude. The RNA sequencing validation is even more limited: 6 patients, with unmatched controls, and the cortical finding hasn’t yet been confirmed histologically. The basal ganglia result also isn’t entirely novel; the role of subcortical circuits in consciousness has been suspected for years, so the AI may be partly rediscovering known biology. And the distance between a circuit-level hypothesis and a bedside treatment is long. Very long, unfortunately.

There’s also the deeper question Pollan’s book kept raising: the AI learned correlates of consciousness, i.e. patterns that distinguish conscious from comatose brains, but that’s not the same as understanding consciousness. The hard problem remains untouched. What the model found is perhaps a map of what breaks. Why there’s something it feels like to be conscious in the first place is a question no dataset, however large, has yet answered.

Why it still matters

All of that said, an approach that doesn’t try to introspect on consciousness but instead learns its neural fingerprints from the outside is new (though I’m sure readers might point me to earlier work!). The bottle has an outside. Whether this particular paper has found the right way to read it, or just one promising way among many, remains to be seen.

What’s your reaction to AI being used to study consciousness? I’d love to hear in the comments.

An uncomfortable history

The DSM as we know it was born in 1980 (the same year I was born). DSM-III wasn’t the first version but is was considered revolutionary by some. Before DSM-III, psychiatric diagnoses in the United States were so inconsistent that two clinicians evaluating the same patient might reach completely different conclusions. The history here is wild. Check out On Being Sane In Insane Places published in Science in 1973 which in some ways sparked the need for DSM-III. (Turns out the paper was perhaps also somewhat fradulent.) The goal of DSM-III was reliability, not validity. I do think for all of its shortcommings, this was a useful concept for many reasons including the reduction of stigma.

Subscribe now

The manual has grown from 106 disorders in DSM-II to 298 in DSM-5, the latest revision. With each new version, some patients have gained access to diagnoses and services, and some patients lost them.

The ASD story: what happens when you redraw the lines

The autism transition from DSM-IV to DSM-5 is a case study I often refer to, partly because of my time as Scientific Director at the Lurie Center for Autism at MGH, and partly because I think it’s a great example of good intention gone wrong. DSM-IV gave clinicians five separate diagnoses under the umbrella of Pervasive Developmental Disorders (PDD): Autistic Disorder, Asperger's Syndrome, PDD-NOS, Childhood Disintegrative Disorder, and Rett's Disorder. DSM-5 collapsed all of them into a single category — Autism Spectrum Disorder (ASD).

The scientific rationale was perhaps defensible. Genetic evidence showed no clear biological boundaries between the subtypes of DSM-IV. That is, the same genes conferred risk across what became known later as the spectrum. That said, its very difficult to untangle intellectual disability from the genetic associations. The spectrum concept seemed to make some biological sense. (The real answer is that we don’t yet know.)

The consequences have been significant. People who had built identity and community around an Asperger’s diagnosis suddenly found themselves without one. Someone who had qualified for services and educational support under DSM-IV no longer met the threshold under DSM-5. And the research literature became harder, much harder(!), to navigate overnight. Studies conducted under DSM-IV criteria couldn’t be cleanly compared to those under DSM-5. We certainly didn’t need the ASD literature to become any more complicated. And yet there it was. Also, the prevalence rise has been, in part, attributed to the change from DSM-IV to DSM-V.

I’ve written about this previously. There is an inherent tension in a document that is simultaneously scientific, clinical, administrative, legal, and social.

The comorbidity problem

Approximately 45% of all diagnosed psychiatric patients meet criteria for more than one DSM category. Nearly half. And for some diagnoses, comorbidity is more common than a single diagnosis.

To be diagnosed with major depression, for example, you need five symptoms from a list of nine, which means two people can share the same diagnosis while having only two symptoms in common. By one estimate, there are 227 unique symptom profiles that meet the threshold for major depression. For PTSD, the number is over 600,000.

Tom Insel, who led the National Institute of Mental Health from 2002 to 2015, called out this problem directly and worked on it while he was in charge. He launched the Research Domain Criteria (RDoC) as an explicit alternative to DSM categories, arguing that DSM diagnoses had no biological validity and that the field needed to start from neuroscience rather than symptom checklists. RDoC never replaced DSM — it couldn’t, practically speaking — but it reframed the debate and inspired a generation of research cutting across diagnostic categories rather than working within them.

The new proposals

Which brings us to today, 2026, and the most serious rethinking of DSM since 1980. The APA's Future DSM Strategic Committee has published a series of papers in the American Journal of Psychiatry outlines a proposed four-domain architecture: (1) contextual factors (social determinants of health, developmental history, quality of life), (2) biomarkers and biological factors, (3) diagnoses, and (4) transdiagnostic features. For the first time, DSM might explicitly build in biological measures.

But I am currently skeptical on the biomarkers bit as someone who spent nearly two decades chasing neuropsych biomarkers.

The chicken-and-egg problem in psychiatric biomarker research is tremendously underappreciated. For decades, the field has defined patient groups using DSM categories and then searched for biological signals within those groups. But if the categories are wrong. If the DSM boundaries don’t reflect actual biological distinctions then the biomarkers you find are just confirming the categories you started with. You’re not discovering biology. You’re finding echoes of your assumptions — the labels you created to have a reliable but perhaps not valid diagnostic category. Breaking that loop requires something radical. For example, we would need large, unbiased studies that measure biology first and let the data tell you where the clusters are.

It’s a tough problem. It’s hard, expensive, and requires phenotyping at a scale the field hasn’t yet achieved. BUT IT CAN BE DONE. It just tasks the courage to do it (and the funding from Congress).

So back to the renaming of DSM… perhaps the best framing is that “Diagnostic and Scientific Manual” is a commitment and a challenge rather than a description of where we are. If so, I’m all for it. Science that names what it wants to be, then does the work to get there, is science I can get behind!

Leave a comment

Subscribe now

The pharmacological wall between neurology and psychiatry is, in my view, substantially more permeable than the institutional one. [I’ll come back to this neurology/psychiatry divide in a future post — it deserves its own essay.] Drugs don’t respect medical specialty boundaries because — surprise — the brain doesn’t organize itself by specialty. Serotonin, GABA, glutamate, the HPA axis — none of these systems stop at the border between a seizure disorder and a mood disorder. The molecules found their way across the corridor because the biology was, of course, already connected.

A paper just published in JAMA Neurology recently, has me thinking deeply about the fundamental overlaps in biology that drive two very different — yet both profound — unmet needs: drug-resistant seizures and suicide.

Share

The main findings of the paper

Barnard and colleagues followed 347 adults with newly diagnosed focal epilepsy for up to six years, tracking psychiatric status at diagnosis and watching to see who became drug-resistant. The question was simple — the best ones in science often are:

Does your psychiatric state at diagnosis predict whether the antiseizure drugs will work?

The answer?! Yes — but not in the way anyone expected. Depression alone didn't predict drug resistance. Anxiety alone didn't predict it. What did — powerfully and independently — was suicidality. Patients with active suicidal ideation or a history of attempts at the time of diagnosis were more than twice as likely to develop drug-resistant epilepsy. I know I put it in bold but stop and let that sink in. More than twice as likely. (We're not squinting at p-values here.) And those with suicidality without a concurrent mood or anxiety disorder showed the strongest effect — nearly three times the baseline risk. Honestly, I'm still processing that finding. [Suicidality without depression predicting resistance more strongly than suicidality with depression tells us this isn't simply about being psychiatrically sicker. Something more biological is going on, but it’s going to take some clever experimental ideas to figure this one out.]

Important implications to think about…

Nearly 40% of this cohort had some psychiatric disturbance at the time of diagnosis and yet fewer than 5% were on any psychotropic medication. Imagine the unmet burden that folks with epilepsy are faced with — can't we do better? I find this to be a major statement on our healthcare system (as if we needed other ways to doubt it) and a signal of how medical specialty boundaries can detract from comprehensive care. Neurologists stop the seizures, psychiatrists deal with mood disorders. Not to be overly reductionist, but I feel there's a bit of "not my job" going on here. [I deeply respect my physician colleagues and believe they are doing their best to care for folks. My point is the system isn't designed to help them do that.]

Seriously — as the authors point out, the Columbia Suicide Severity Rating Scale takes under fifteen minutes to administer and is easy to implement. A positive screen identifies patients at nearly three times the baseline risk of drug resistance — that’s very actionable in my mind. It could change how aggressively a clinician escalates treatment, how quickly they consider surgery.

Beyond the immediate clinical implication, this finding points to something else that I think is important. If suicidality is a marker of a brain where serotonin, glutamate, and inflammatory systems are already dysregulated — and those are exactly the systems that drive seizure resistance — then new antiseizure medications developed for treatment-resistant epilepsy may carry psychiatric benefits that nobody has thought to look for yet. The next lamotrigine might already be in a trial, quietly doing something important in depression or anxiety that the field hasn’t noticed. That’s the promise of taking intersectional biology seriously. It’s a good reminder to avoid “staying in your lane”.

Leave a comment

I’m actually a believer in mindset as a powerful healer. Not woo (i.e. I’ve written about science meets woo). Not wishful thinking. Real, measurable biological change. The biology is driven by expectation, context, and the extraordinary machinery the brain deploys when it believes something is happening. The phenomenon we call “placebo” is probably not one thing. It’s likely a collection of biological mechanisms. We’ve collapsed a lot of mechanisms under a single name because they share the common feature that no “drug” caused them.

That’s what makes placebo so fascinating and so deeply inconvenient, simultaneously awesome and infuriating.

In clinical trials for major depressive disorder, placebo response rates typically run between 35 and 50 percent. That means that in trials designed to test whether a drug works, roughly four in ten people assigned to the “placebo” get significantly better. And the number has been climbing over time. A landmark 2002 paper in JAMA — titled, with unusual bluntness, “Placebo response in studies of major depression: variable, substantial, and growing” — documented this trend.

The practical consequence is that as placebo response rises, the gap between drug and placebo narrows. This also means that a drug that convincingly beat placebo in a 1990 trial might fail the identical trial today — not because the drug got weaker, but because the placebo got stronger. My colleagues Maurizio Fava and Diego Pizzagalli, along with Ted Kaptchuk at BIDMC — who has spent more of his career studying placebo than almost anyone alive — are co-authors on a paper I was part of that tries to get at the neurobiological underpinnings of exactly this problem. The challenge is real, it is growing, and it is not well understood.

This is not fake. It’s in the brain.

PET imaging studies have shown that placebo administration produces dopamine release in the nucleus accumbens — the brain’s reward hub — in proportion to the expectation of improvement. The brain, in other words, can generate its own pharmacology when it expects relief.

This connects to the concept of the predictive brain (indeed your brain is lying to you!). The brain doesn’t passively register the world. It constantly generates predictions about what’s coming and updates them based on incoming signals. In that framework, placebo isn’t mysterious at all. It’s the brain’s prediction of improvement — “I took a treatment, therefore I should feel better” — generating a real biological cascade. The placebo is the prediction.

It's a reminder for me that the boundary between mind (and/or brain) and body was always more porous than medicine liked to admit — something Bessel van der Kolk has spent a career documenting in a different but deeply related context. See the brilliant book: The Body Keeps Score.

Which means placebo response isn’t noise contaminating our signal. It’s the brain’s own therapeutic machinery, activated by context and expectation. And we’ve been designing trials to minimize it rather than understand and exploit it.

Why is the placebo getting stronger?

Short answer is, “I don’t know”. Several theories compete and they’re probably all partially right and mostly wrong.

The first is purely statistical. It’s called regression to the mean. People enter trials at their worst, and some proportion would have improved regardless of treatment. As trial enrollment has expanded and inclusion criteria broadened, more people who were going to naturally improve are ending up in placebo arms.

The second is the therapeutic ritual effect. Participants in clinical trials receive something that most patients in regular care do not. They get close monitoring, frequent visits with attentive clinicians. This creates a the sense that someone is watching and that their improvement matters. Ted Kaptchuk’s work has shown that the ritual of receiving care — the white coat, the careful attention, the act of being treated — is itself biologically active. The placebo arm in a well-run trial is not actually receiving nothing. It’s receiving a highly optimized therapeutic (ritual) encounter.

The third, and to me the most interesting, is the information environment. Patients now arrive at clinical trials shaped by decades of direct-to-consumer pharmaceutical advertising, health media, online patient communities, and a broader cultural narrative that treatment works. Expectation is primed before they walk through the door. When someone enrolls in a depression trial in 2025, they bring with them an accumulated belief — biological in its expression — that taking a pill as part of a clinical trial is associated with getting better. That belief is not inert.

Subscribe now

The double-blind is kind of a fiction for psychoactive drugs

For most psychoactive drugs, the double-blind trial is not actually blind.

Consider caffeine — a common, pschoactive drug. If I gave you a capsule and told you it might be caffeine or placebo, could you tell? Of course you could — within twenty minutes. (OK - we’ll maybe not if you drink a tremendous amount of coffee every day, but think about being caffeine naiive for a while.) Now extend that logic to antidepressants with distinctive brain activity and side effects. To opioids... (good luck masking those!) To ketamine… (clearly not placebo here!) To psychedelics, where the subjective experience is so profound that essentially no participant is genuinely uncertain about their assignment even with “active comparators”.

This matters so much in a trial. When a participant knows — or strongly suspects — they’re on the active drug, their expectations shift accordingly. And when a participant suspects they’re on placebo, their expectations shift the other way. You are no longer measuring drug versus placebo. You are measuring drug-plus-heightened-expectation versus placebo-plus-dampened-expectation. The effect size you report is a blend of pharmacology and psychology that you cannot cleanly separate.

A striking example from outside psychiatry. TauRx’s LMTM, a methylene blue derivative tested in Alzheimer’s disease, colors urine greenish-blue. The solution to maintain blinding was to give the placebo arm a low dose of the active compound — just enough to produce the same urine color, but presumably too low to have a biological effect. This was an ingenious and heroic trial design. It is also, in retrospect, revealing… the “inactive” placebo dose later showed signs of possible biological activity, which threw the entire interpretation of the trial into question. When you have to give your placebo arm a dose of the drug to maintain the fiction of blinding, you have already left the world of the clean double-blind. This is an extreme case, but the underlying problem — that many psychoactive compounds announce themselves to the trial participant — is not. This is one of the most serious and underacknowledged methodological challenges in psychiatric drug development, and it’s only going to get harder as we develop more potent — more subjectively psychoactive compounds.

What we should do — both things

Managing placebo in clinical trials is not optional. Stricter enrollment, run-in periods, careful site selection — these are essential, and anyone running a serious CNS trial is working hard on exactly this. At Sensorium we think about it constantly.

But managing placebo and understanding placebo are not the same project.

One underappreciated angle is durability. Placebo responses, while real, tend to decay. Drug effects — when genuine — persist and often deepen. That temporal signature is one of the clearest distinctions between the two, and one of the better arguments that the drugs we’re developing are doing something the brain cannot sustain on its own.

The deeper questions remain wide open. What makes some people placebo responders and others not? We know susceptibility varies meaningfully across individuals — some people are reliably stronger placebo responders than others, and that variability appears to have neurobiological and genetic underpinnings worth understanding. Can we predict it from baseline neurobiology and use that prediction to personalize treatment?

Placebo is not a methodological problem to be engineered away. It’s a signal about what the brain is capable of. The work is to understand it well enough to use it wisely — and to make sure the drugs we develop are doing something the brain cannot do for itself.

What’s your experience with placebo — as a patient, a clinician, or a researcher? I’d love to hear in the comments.

Leave a comment

Subscribe now

In lectures, complex molecular structures came off the blackboard at a pace that assumed you already knew how to see them. Drawing molecules is a physical skill — every bond angle matters, every line has meaning, a single misplaced atom changes the identity of the compound entirely. You develop a hand for it, the way a musician develops an ear.

As an undergraduate at NC State, I had loved organic chemistry for its logic — the rule set, the framework, the way a synthesis could be planned like an argument. I understood the moves. But at Berkeley I started to see something else entirely. Molecules have architecture. Not metaphorically — literally. They have key elements, elegant spans, forms that solve structural problems with the same ingenuity as a great building. Some parts are modest and efficient. Others are incomprehensibly elaborate, as if nature were showing off. I stopped trying to decode the diagrams and started seeing them as objects — three-dimensional, spatial, sometimes breathtakingly beautiful.

That shift — from logic to beauty — is what this essay is about. [I wrote about the intersection of science and aesthetics in an earlier post — the idea that beauty and rigor aren’t opposites but partners. Molecules are where I feel that most acutely.]

Strychnine is one of those molecules that is simply beautiful.

A molecule that shouldn’t exist

If you’ve heard of strychnine, you probably know it as a poison — the villain’s weapon of choice in Agatha Christie novels. [The Mysterious Affair at Styles 1920]. That reputation is accurate. Strychnine is extraordinarily toxic in small doses, acting as a powerful antagonist at glycine receptors in the spinal cord. [Glycine is one of the brain’s principal inhibitory neurotransmitters — think of it as a brake on motor neurons. Block those brakes and you get violent, uncontrolled muscle contractions. That’s what strychnine does.] But forget the poison for a moment. Look at the structure.

Draw strychnine on a page and even to an untrained eye it looks formidable — a dense tangle of rings that seems almost combative. Look at it in three dimensions and you start to appreciate what nature built: seven contiguous stereocenters [seven consecutive carbon atoms, each locked in a precise spatial orientation that cannot be accidentally swapped], two nitrogen atoms tucked into a cage-like architecture, five rings fused in a configuration that chemists spent decades trying to understand, let alone replicate. It looks, as I once heard in that Berkeley classroom, like a problem that shouldn’t have a solution.

In 1954, R.B. Woodward published the first total synthesis of strychnine — building it from scratch, step by step, from simple starting materials. [Total synthesis is the chemical equivalent of reconstructing a cathedral from sand and gravel, with no blueprint except the building itself.] It remains one of the greatest achievements in the history of organic chemistry. Woodward did this without computational modeling, without modern synthetic methods — and without fully understanding what the molecule actually did in the body. He was drawn to it because it was beautiful and seemed impossible. Those two things together were irresistible. I understand that completely.

The older scientists

Strychnine comes from the seeds of Strychnos nux-vomica, native to South Asia, and related species across tropical Africa and Southeast Asia. Long before any Western chemist isolated the compound — before the word alkaloid even existed — people knew exactly what these seeds could do. Traditional practitioners in South Asia used carefully calibrated doses in Ayurvedic medicine for centuries. In parts of West Africa, related Strychnos species featured in arrow poisons and ritual contexts. That knowledge was accumulated through the most patient and courageous form of empirical science imaginable: direct human observation, across generations, with very high stakes for getting it wrong.

This is what ethnobotany actually is. Not the romanticization of indigenous knowledge, and not its dismissal — but the serious scientific recognition that when a healer has used a particular plant for a particular purpose across ten generations, that is a hypothesis that has survived ten generations of empirical testing. The molecular logic wasn’t articulated in the language of receptors and binding affinities. But it was tested, refined, and transmitted with extraordinary care. [I think about this every time I read a clinical trial that “discovers” a therapeutic effect an indigenous community has known about for centuries. We should be more humble about who the real pioneers are.]

Ethnobotany sits at the intersection of anthropology, ecology, chemistry, and medicine. Its field is the living world — forests, markets, ceremonies, the memory of elders. Its laboratory is wherever that knowledge meets modern chemistry. And its central insight is simple: nature has been running experiments on the human nervous system for far longer than we have.

What we got wrong

For the past few decades, drug discovery leaned heavily on high-throughput screening — testing enormous libraries of synthetic compounds against biological targets at industrial scale, looking for hits. It is an impressive technological achievement. In the CNS space especially, it has produced a fairly disappointing return. [Roughly 85% of CNS drugs in use today trace their origins to natural products or natural product-inspired scaffolds — a number that still surprises people.] High-throughput screening treated natural products largely as noise: too complex, too hard to synthesize, too difficult to optimize. The field wanted clean, flat, drug-like molecules. Nature kept offering strychnines. [To be fair, there are many gems out there with complexity somewhere between synthetic libraries and strychnine.]

What we missed is that the complexity is the point. Plants don’t make molecules like strychnine by accident. The elaborate architecture, the precise stereochemistry, the specific receptor affinity — these are the products of hundreds of millions of years of evolutionary pressure. Plants can’t run. They can’t fight. Their chemistry is their immune system, their communication network, their memory. When a plant produces a molecule that binds exquisitely to a receptor in the human nervous system, that is not coincidence. It is the result of an evolutionary conversation between chemical systems running far longer than our species has existed. [This connects directly to something I wrote about earlier — chemical co-evolution between brains and nature’s molecules. Strychnine and the glycine receptor are a striking example: a plant evolved a compound that precisely targets one of the oldest inhibitory systems in the vertebrate nervous system. That precision deserves to be studied, not screened out.]

The signal was always there. We built machines to generate more noise.

What it means to actually listen

Taking ethnobotany seriously means treating traditional knowledge as what it is: a hypothesis-generating system of extraordinary depth, earned through observation across time. It means going into the field — literally, into forests, communities, and herbaria — and listening before extracting. It means protecting those forests, because a molecule we haven’t found yet doesn’t cease to exist just because we haven’t looked. And it means asking not just what a plant makes, but why — what does the evolutionary logic tell us about the biological systems these molecules were shaped to interact with?

If you want to go deeper, start with Mark Plotkin — ethnobotanist, Amazon explorer, and one of the most compelling voices on why this work is urgent. His book Tales of a Shaman’s Apprentice is a masterclass in what it looks like to actually listen, and his podcast Plants of the Gods (co-hosted with his mentor’s legacy in mind) is a great entry point. Wade Davis, another student of the legendary Harvard ethnobotanist Richard Evans Schultes, brings a similar depth — The Serpent and the Rainbow and One River are both worth your time. These are people who have spent careers in the field doing exactly what I’m arguing for: treating indigenous knowledge not as folklore to be mined but as science to be learned from.

How many molecules like strychnine are still out there, in roots and bark and soil we haven’t listened to yet? The answer depends less on our technology than on our willingness to protect what remains, slow down, and take seriously the signals that have been there all along.

What natural product or story of medical discovery has captured your imagination? Tell me in the comments.

Leave a comment

Leave a comment

What Was the Leucovorin Story — and Why Does the Retraction Matter?

For readers who haven’t followed this closely, leucovorin (also called folinic acid) is a biologically active form of folate that has been explored as a potential treatment for a subset of autistic individuals — particularly those with disruptions in folate transport or so-called folate-receptor autoantibodies. The scientific idea wasn’t unreasonable: if certain metabolic pathways affecting brain development are altered, could supporting those pathways improve communication or cognitive function? A handful of small studies over the past decade suggested possible benefits, which led to growing excitement — and, in some corners of the internet, over-interpretation.

The recently retracted trial had been described as one of the largest and most promising investigations of leucovorin in autism. How did it play out? Concerns emerged about the integrity of parts of the data and analysis, ultimately leading the journal to withdraw the paper after a reanalysis could not replicate the outcome statistics.

What makes this moment complicated is that the leucovorin hypothesis itself isn’t inherently “wrong.” There remains legitimate interest in folate biology, immune signaling, and metabolic diversity within autism. (See my previous post on the topic.) But a single study — especially one that becomes amplified in public conversations — can shape expectations in ways that outpace the evidence. When that study falls apart, the emotional fallout can ripple far beyond the lab.

Here’s the retracted paper: https://link.springer.com/article/10.1007/s00431-024-05762-6

This moment also lands in a complicated cultural landscape. (Complicated is the most generous term I could come up with.) Over the past decade, autism research has increasingly been pulled into broader debates about causation — from vaccines to medications — debates that often move faster than the data. When studies are overstated or misinterpreted, the consequences don’t stay confined to academic journals. We’ve already seen how fragile public trust in biomedical science can become, and how misinformation (Tylenol being a prime example) can ripple outward into real-world behaviors. Most concerning is that in recent years, declining vaccination rates in some communities have coincided with resurging measles outbreaks across parts of the United States — a reminder that how we communicate science matters as much as the science itself. (If you don’t know the history of the vaccine science in ASD, I’ll point you here and here as a start.)

Subscribe now

I don’t see this as a reason to stop exploring new ideas. If anything, it’s a reminder of how careful we need to be with early signals — to stay curious without mistaking momentum for certainty.

What I come back to is the notion that optimism in science is a choice. Not blind belief in any single study, but a decision to trust the process — the corrections, the re-thinking, the slow refinement that moves us closer to truth. I choose optimism because I have worked side-by-side with folks that work unbelievably long hours, year after year with relatively modest pay to chase scientific ideas that could have impact for people with ASD. I choose optimism because science keeps learning, because better tools keep emerging (and at an accelerated pace), and because every honest course-correction makes the next discovery more reliable for the people who need it most.

I hope you will choose optimism too.

Leave a comment

Revisiting that framework got me wondering how much of this intuition is actually supported by data. Trust, after all, sits at the base of Lencioni’s pyramid in the book—but trust is a slippery thing to measure in my opinion. A relatively recent paper that I came across in Applied Psychology approaches the concept from a different angle, looking not at trust directly, but at the values people bring into teams. It turns out that teams characterized by benevolence-oriented values—concern for others, fairness, and cooperation—consistently outperform those without them. These factors when combined with vulnerability are the recipe I favor for trust.

Subscribe now

The result from this paper is not an outlier. Decades of organizational research point to the same underlying pattern: teams function better when members assume good intent in one another. Large meta-analyses in the Journal of Applied Psychology have shown that intrateam trust is robustly associated with higher performance, stronger commitment, and better coordination—even after controlling for leadership quality and prior success. More recent syntheses in the Annual Review of Organizational Psychology and Organizational Behavior argue that trust works less like a “soft” cultural nice-to-have and more like an enabling condition: it lowers the cognitive and emotional cost of working together under uncertainty. When trust is present, teams can disagree without fragmenting, commit without full information, and adapt without blame. From my experience, the ability to disagree and push one another in science without fragmenting or shutting down is what drives teams to impactful work.

What’s especially interesting is that this story doesn’t end at surveys and performance metrics — not that those aren’t important but the neuroscientist in me wants more! Over the last decade, neuroscientists have started asking whether trust and coordination leave measurable signatures in the brain itself. Using techniques like hyperscanning fMRI and EEG—where multiple brains are recorded simultaneously—researchers have shown that teams who cooperate effectively under uncertainty exhibit synchronized neural activity across regions involved in attention, prediction, and social inference. In other words, high-functioning teams don’t just behave differently; they appear to think differently together. Trust, it turns out, may be less a feeling than a shared neural state that makes collective prediction possible.

EEG hyperscanning has found that groups performing joint decision-making tasks under uncertainty exhibit higher inter-brain synchrony during cooperation than during independent or competitive activity, and that the degree of synchrony correlates with measures of trust, shared attention, and adaptive performance. Separate work on speaker–listener neural coupling using fNIRS has demonstrated that during effective communication, the neural responses of interacting individuals become aligned over time, and that stronger coupling predicts better information transfer and group outcomes—further supporting the idea that aligned neural dynamics accompany successful interpersonal coordination.

What is hyperscanning?
Hyperscanning refers to simultaneously measuring brain activity from two or more people while they interact. Instead of asking how a single brain responds to a task, hyperscanning asks how brains couple during real collaboration—revealing neural synchrony that emerges only when people work together.

What ties all of this together is the degree of uncertainty. In environments where outcomes are well defined, teams can rely on process, hierarchy, and checklists. Biotech is not one of those environments. When uncertainty is high and information is incomplete, teams are forced to rely on one another’s judgments, intuitions, and partial signals. In that context, trust becomes more—not less—important. It lowers the friction of collaboration, allowing disagreement without defensiveness and rapid iteration without blame. The emerging neuroscience fits this view: inter-brain synchrony appears most strongly when people are actively coordinating under uncertainty, suggesting that trust may function as a biological enabler of shared prediction rather than simply a cultural value.

If trust is a shared cognitive state rather than a feeling, it raises a different question: how often do we actually design teams to support that kind of thinking together?

Leave a comment

Pause for a minute and think about that.

Anthropologists like to point out that languages don’t simply give names to things that already exist; they carve the world into categories and teach us what to notice. Of course what we notice depends on the context of where and how we live. There are many examples but given the Boston winter and the white ground around me, I’ll focus on the simple word, “snow”. To me snow matters in simple binary ways — will my kids have school? Will there be enough snow to go sledding? In some cultures, there are dozens of words that we would interpret/translate to be “snow”. The popular (if oversimplified) claim is that because snow matters so much in snowy environments, speakers of those languages developed rich vocabulary to distinguish wet snow from drifting snow, crusted snow from drifting snow, snow that’s good for sled runners vs. snow that’s not.

Subscribe now

The vocabulary we inherit and use teaches us which distinctions are important in our world. Vocabulary also deliberately and unintentional evolves!

Once a language gives a name to something, it becomes easier to notice, to remember, to argue about, and ultimately to think about in more nuanced ways.

So when we ask, what is anxiety? — the answer is not so simple. This got me quite curious about the evolution of the concept given that anxiety is at once the most intimate and the most alien of human experiences. Anxiety is the background signal of a predictive brain—the constant scanning for what might come next. It is the cost (literally a metabolic cost) of imagining futures before they arrive. Anxiety is the price of foresight. To map the evolution of anxiety is not merely to trace the history of a medical diagnosis; it’s to follow the history of how humans have understood themselves. Whether framed as the “trembling of the heart” in the Nile Valley, the assault of a Mesopotamian ghost, the depletion of “nervous force” in the industrial age, or the firing of a “false suffocation alarm” in a modern neuroimaging lab, the phenomenon of anxiety remains a constant somatic and psychic reality, relentlessly re-interpreted through the lens of the prevailing zeitgeist.

I became curious about the history of the concept of anxiety and found this paper that’s a good start if you’re curious. Cicero, in his Tusculan Disputations, provides one of the earliest technical distinctions in the lexicon of anxiety, anticipating modern trait/state theory. He distinguishes between: (1) Angor: A transitory outburst of distress, often associated with physical suffocation or panic. This aligns with the modern “Panic Attack.”; (2) Anxietas: A constitutional proneness to worry; a lasting state of mind. This aligns with the modern “Generalized Anxiety Disorder” or “Neuroticism” (Trait Anxiety).

Anxiety Before the Mind

In the ancient world, anxiety wasn’t a mental disorder—it was something that happened to the body. This is certainly how I experience anxiety. In Mesopotamia and Egypt, fear, panic, and distress were felt in the heart and chest, not the brain, and often explained as invasions by external forces. For me it’s very much a body/heart feeling. Babylonian texts describe people who were constantly frightened, sweating, short-tempered, and oppressed in the heart—symptoms we’d instantly recognize today. The cause, however, was supernatural: the “hand of a ghost” or night demons that pinned victims down in terror, descriptions that eerily resemble panic attacks and sleep paralysis. Treatment mixed herbs with ritual exorcism. Anxiety was something that entered you.

Why Anxiety Feels Everywhere Now

Anxiety has always been part of the human condition, but it is showing up more often and in more people than it once did. In the U.S. today, about 18–19% of adults report anxiety symptoms in any given two-week period and nearly 1 in 5 adults will meet criteria for an anxiety disorder in a year.

That’s not just a medical problem; it’s a consequence of a world that asks our brains to predict more and with more uncertainty than ever before. The same system that once helped us survive real, immediate danger now scans headlines, economy forecasts, social media feeds, and imagined futures. If anxiety reflects the cost of foresight, we need new tools to help tune that system, not just dampen it. Revisiting botanical strategies from the past informed by modern neuroscience may be one avenue worth exploring, not as nostalgic throwbacks but as biologically grounded, mechanism-driven interventions that restore balance to bodies and minds overwhelmed by prediction.

If language shapes what we notice, how would you name and make sense of your own anxiety? What words resonate for you?

Leave a comment