The Trip Was the Side Effect
The Heresy at the Finish Line
Sometime in the next year, a drug the United States banned as dangerous and worthless in 1970 is likely to become prescription medicine. It is psilocybin, the molecule in magic mushrooms. Two large trials have now shown that one or two doses, taken under supervision, can lift severe depression that no ordinary antidepressant had touched, and hold it down for weeks. The company carrying a synthetic version through the Food and Drug Administration expects to finish its application by the end of 2026. That would put a decision in late 2026 or early 2027. Ketamine’s chemical cousin already reaches patients, but ketamine is not a classic psychedelic. Psilocybin would be the first of that older, trip-producing family, the one that includes LSD, to be stamped as medicine.
But the more interesting story is not the approval. It is a quiet shift in what these drugs are understood to be. When the psychiatrist Humphry Osmond coined the word “psychedelic” in 1957, he built it from Greek for “mind-manifesting.” The drug’s job, in that telling, was to reveal the hidden contents of the mind, and the healing was thought to live inside the experience. A lab at the University of California, Davis has proposed something blunter. David Olson, a chemist who founded the university’s Institute for Psychedelics and Neurotherapeutics, argues that what these molecules mainly do is physical. They push brain cells to regrow the connections that wither in depression. Olson gave the property its own name, psychoplastogen, a molecule that drives the brain’s built-in capacity to rewire itself. The heretical claim buried in that word is that the trip is not the medicine, but a side effect. If the healing is structural repair and not revelation, the hours of hallucination on the couch are not the cure. They are something to engineer away.
That reframe turns a therapy movement into a drug race, and the race has already split in two. On one side is the molecule about to win approval, and it kept the trip. A Compass Pathways patient still swallows a dose and spends six supervised hours, sometimes overwhelmed, under eyeshades. On the other side is a fast-growing camp betting the trip is disposable. AbbVie has put up to two billion dollars behind one company’s non-hallucinating versions. Olson’s own startup and several rivals have raised hundreds of millions more to build psychedelics that heal without ever altering perception. One side is first to market with the experience intact. The other is wagering that the experience was never the point.
So the whole thing turns on a single question. Is the trip the medicine, or the side effect? Everything else hangs off the answer. Can these drugs scale past the eight-hour session? Can a system built for daily pills absorb them? How far past depression does the same trick reach? In a mouse, the experience and the repair already come apart. A redesigned molecule can regrow the wiring and produce no trip. By the end of the decade we will know whether the trip can be cut away and the cure kept, or whether the trip was the point.
From Bicycle Day to the Plasticity Pivot
On April 16, 1943, a chemist named Albert Hofmann was working late at the Sandoz pharmaceutical laboratories in Basel, Switzerland. He was resynthesizing a compound he had first made five years earlier, then shelved. He had been hunting for a circulatory stimulant among the derivatives of ergot, a fungus that grows on rye. A trace of the substance reached his skin or his lips. The afternoon turned into a stream of vivid, shifting images. Suspecting the compound, lysergic acid diethylamide, Hofmann did the thing no cautious chemist would advise. Three days later, on April 19, he swallowed a deliberate dose of 250 micrograms, a dose he assumed was trivially small. It was a heavy dose, many times the few dozen micrograms at which the drug’s effects begin. The effects came on so hard that he asked his lab assistant to take him home. Cars were restricted in wartime, so the two of them rode home by bicycle. The ride was terrifying and full of hallucinations, and it became the founding legend of the field. People still mark April 19 as Bicycle Day. Hofmann had not invented a party drug. He had stumbled onto a molecule active at doses measured in millionths of a gram. That potency meant it was acting on some precise piece of the brain’s machinery, not flooding the whole thing.
Sandoz saw a research tool and a possible psychiatric medicine. The company began shipping the compound to scientists and clinicians around the world under the brand name Delysid. It invited them to find uses for it. They did, in volume. By the mid-1960s researchers had published more than a thousand clinical papers, describing the drug given to roughly forty thousand patients. There were also dozens of books and a series of international conferences. One leading approach was called psycholytic therapy. It used repeated low-to-moderate doses to loosen a patient’s defenses during ordinary talk therapy. By the mid-1960s more than thirty European clinics practiced it. Trials tested the drug for alcoholism, for depression and anxiety, and for the dread of dying in terminal cancer patients. For a while it was treated as the coming revolution in psychiatry. The hope was that it might replace electroshock and the surgical cutting of brain tissue then in use. The work was uneven by modern standards, often missing control groups, but it was serious, funded, and mainstream.
Then it collapsed. The same molecules escaped the clinic into the 1960s counterculture, became badges of youth rebellion, and caused a political backlash that drew no line between a research compound and a street drug. In 1970 the United States Congress passed the Controlled Substances Act. LSD, psilocybin, and their relatives were placed in Schedule I, the most restrictive tier. These were defined as drugs with high abuse potential and, by law, no accepted medical use. The classification did more than regulate. It made the drugs nearly impossible to obtain, fund, or study, and it turned simple possession into a felony. Promising lines of research were abandoned half-finished, careers redirected, supplies destroyed. For roughly a generation, serious clinical study of psychedelics in humans effectively stopped. The thousand papers already in print sat as historical curiosities, not foundations to build on.
The thaw, when it finally came, was deliberately cautious. In 2006 a team at Johns Hopkins University published a rigorous double-blind trial of psilocybin in healthy volunteers. The team was led by Roland Griffiths, a veteran drug researcher who had spent decades studying caffeine and how dependence works. The design was built to satisfy skeptics. Participants received either psilocybin or a stimulant comparison drug, without being told which. They were screened in advance and supervised throughout, in a calm room, often lying down with eyeshades and music. Two-thirds of those who took psilocybin later rated the session among the five most meaningful experiences of their lives. The lift in mood and outlook held for months. Coming from a respected lab and a scientist with no countercultural baggage, the paper did what a flashier study could not. It made psychedelic science respectable again, and reopened the funding and regulatory doors that had been shut for thirty years.
A second effort had been grinding forward for years. The Multidisciplinary Association for Psychedelic Studies, a nonprofit founded in 1986, had been quietly funding trials of MDMA, the compound sold on the street as ecstasy. The target was post-traumatic stress disorder. Patients took the drug only a handful of times, alongside intensive psychotherapy. The early results were strong enough to win the program a breakthrough designation from regulators. Over the next decade many small trials piled up across the field. They tested psilocybin for depression, for alcohol and tobacco addiction, and again for the anxiety of terminal illness. A consistent, strange signal ran through all of them. One or two doses, not a pill swallowed every day, could produce drops in symptoms that lasted weeks or months. The MDMA program itself later stumbled. In 2024 the Food and Drug Administration declined to approve it. It faulted the commercial arm, Lykos Therapeutics, for trials whose blinding and durability it did not trust. But the underlying pattern, lasting benefit from a few exposures, kept surfacing across different drugs and different diagnoses.
Proof that the category could become ordinary, reimbursed medicine arrived from a chemical cousin. Ketamine is an anesthetic in use since the 1960s, a dissociative that warps perception without producing classic hallucinations. It turned out to lift severe depression within hours. Johnson & Johnson isolated one mirror-image half of the ketamine molecule, named it esketamine, and packaged it as a nasal spray called Spravato. The FDA approved it in March 2019 for treatment-resistant depression, the kind that has failed at least two standard antidepressants. It was to be used alongside an oral pill. In January 2025 the agency cleared Spravato to work on its own. That made it the first and only standalone drug approved specifically for that condition. It is given inside a certified clinic. The patient is watched for about two hours while the sedation and dissociation pass, then sent home. That supervised, in-clinic delivery, paid for by insurance, became the concrete anchor for everything the psychedelic field hoped to follow. It was a mind-altering compound the medical system had agreed to stock, monitor, and reimburse.
But the anchor sharpened a question no one could answer. Standard antidepressants, the Prozac family that has dominated psychiatry since the late 1980s, rest on a tidy story. Depression was thought to come from a shortage of serotonin, one of the chemicals that brain cells use to talk to one another. The drugs work by raising the amount of serotonin that lingers in the gaps between cells. That story has a built-in catch. The pill has to be taken every day. If a patient stops the pill, the chemistry reverts within weeks and the depression comes back. That is exactly how a shortage should behave. But a single dose of psilocybin, or a short course of esketamine, produced relief that outlasted the drug in the body by weeks or months. A shortage that must be topped up daily cannot explain a benefit that persists long after the molecule has cleared. Something other than moment-to-moment chemistry was changing, and it was staying changed.
The answer turned out to be visible under a microscope. Nerve cells connect to one another through countless tiny contact points. In depression those connections thin out and pull back. Researchers at Yale had shown around 2010 that ketamine could rapidly regrow them. In 2018, David Olson’s lab at UC Davis ran the obvious next experiment. They grew nerve cells from the brain’s cortex in a dish. Then they exposed the cells to LSD, to DMT (the active ingredient in the Amazonian brew ayahuasca), and to ketamine, and watched what happened. Every one of them sprouted new branches. The branches grew fresh dendritic spines, the tiny nubs where one nerve cell wires itself to a neighbor. Several of the classic psychedelics did it more powerfully than ketamine. The effect was structural. It was real physical rewiring, not a passing chemical nudge, and it stayed after the drug was gone. This was the discovery behind the term psychoplastogen. The durable benefit now had a physical cause. The supervised hours of hallucination began to look incidental. And a question that had belonged to therapists and seekers passed to neuroscientists, and to the companies that would race to engineer it.
The Trip and the Repair
The structural account of mental illness makes a claim that sounds almost too physical to be about the mind. Depression, post-traumatic stress disorder, and addiction are not, at root, the wrong amount of some chemical sloshing between cells. They are a problem of wiring. The prefrontal cortex is the part of the brain behind the forehead, the one that handles planning, judgment, and mood. Under chronic stress, its neurons begin to pull in their branches. The connections retract, and the circuits those connections built go quiet. Imaging of people with long-running depression shows that same region measurably smaller. This structural view is not only about depression. Addiction is different. Its reward circuit is worn so deep that it overrides everything around it. Trauma fixes a fear memory too strongly and will not let it fade. The common thread in all three is a brain that has lost the flexibility to rewire itself. It is frozen in a shape that no longer serves it. Repairing the wiring and restoring the flexibility should ease the symptoms.
The reason to take that idea seriously is ketamine. The connections it regrows are not just a sign of recovery, they are its cause. When researchers used a tool to erase the fresh growth ketamine had produced, the animal’s depression-like behavior came back, even though the drug had already done its work. That causal link overturned an old assumption: that rebuilding wiring in an adult brain is slow, maybe impossible. The trouble is that ketamine is too crude a tool. It is dissociative, which means it makes a person feel detached from their own body and surroundings. It can be abused as a recreational drug, and heavy users can grow dependent on it. And it is harsh on the bladder at the high, repeated doses some patients need. The useful question became whether the same fast regrowth could be triggered by something cleaner and more precise. The search led to a single receptor.
Psilocybin, LSD, and DMT grow in different mushrooms and plants. On the page they look unrelated. Yet they funnel down to the same piece of hardware in the cell. A receptor is a protein lodged in the cell’s outer wall. It behaves like a lock, and only molecules of the right shape are the keys that fit it. All three drugs fit the same lock, a receptor that normally answers to serotonin and carries the label 5-HT2A. It is concentrated in the cortex, the brain’s outer thinking layer. That is part of why these drugs act on perception and thought, not on something like heartbeat. The evidence that this lock is the one that matters is clean. When a second drug blocks the receptor, no amount of psilocybin produces a trip. Without that blocker, the visions arrive. So the 2A receptor is where hallucination begins. For years that seemed to settle the question. The receptor that produces the trip is the receptor the drugs act on. The complication came from trying to measure a trip in an animal that cannot report one.
You cannot ask a mouse whether the walls are breathing. You can watch its head. A mouse given a classic psychedelic does a fast, involuntary side-to-side flick. This is the head-twitch response. How hard the mouse twitches tracks how strongly the same molecule makes humans hallucinate. It became the field’s cheap, reliable substitute for the trip, the readout a chemist measures when testing a new compound. More twitching means more trip. The twitch itself runs through the 2A receptor. Mice that lack that receptor do not flick. That ties the behavior to the very lock the drugs open. So the chain looked complete. A drug keys into 2A, the mouse twitches, the human hallucinates. Then the chain broke.
Some molecules key into the 2A receptor, switch it on by every chemical measurement, and produce no twitch and no hallucination. One of the best studied is lisuride, a drug long prescribed for Parkinson’s disease and migraine. It grips the same receptor as LSD, and patients taking it do not trip. That should be impossible if the receptor were a simple on-off switch. The way out is an idea called biased signaling. A receptor is not one button wired to one outcome. Think of it as a doorbell that rings bells in several different rooms inside the cell. The exact shape of the key that turns it decides which bells ring. One key rings the bell in every room. Another rings just one. LSD rings the bell for hallucination and, it seems, the bell for repair. Lisuride rings a different set, and leaves the hallucination bell silent. If those bells can be rung one at a time, then somewhere there should be a key that rings only the repair bell. That turns “non-hallucinogenic psychedelic,” which sounds like a contradiction in terms, into a real design target. A molecule that healed without the trip would be worth an enormous amount.
The most striking clue to where that key might come from landed in 2023. A team at UC Davis led by Maxemiliano Vargas, a graduate researcher in Olson’s lab, pressed on a question almost too simple to ask. The body bathes its own brain in serotonin, and that serotonin keys into the 2A receptor all day long. Why does the body’s own serotonin never trigger hallucination or growth, when the psychedelics that hit the very same receptor do both? The answer was geography. Not all of a cell’s 2A receptors sit on the outer wall facing the world. A large share sit inside the neuron, on the membranes of compartments deep within it. And it is those interior receptors, not the ones on the surface, whose activation grows new connections. The key detail is this. Serotonin is a water-loving molecule, and it carries an electric charge. The cell wall is oily. So serotonin cannot pass through it. It is locked outside, and can ring only the receptors on the surface. Psychedelics are oily, fat-loving molecules. They slip straight through the wall to reach the locks held inside, the ones that build. Whether a compound can rewire a brain comes down to whether it can cross the wall to those interior receptors. The difference is geography, not magic.
The team then ran the experiment that turns a nice story into evidence. They gave a patch of neurons a molecular pump, one that hauls serotonin across the wall and into the cell’s interior. Then they supplied serotonin. Handed a way in, the body’s own serotonin began doing exactly what the psychedelics do. It grew new connections, and the mice carrying those altered neurons started to head-twitch. The same molecule that does nothing from outside the cell now produced both the repair and the telltale sign of a trip. All it needed was a way to reach the locks inside. That nearly settles it: location, not the identity of the molecule, is what counts. It also raised a stranger thought. Serotonin may not even be the signal those interior receptors were built to receive. The body may use some other, oilier messenger to reach them, one we have not identified yet.
Switching on an interior receptor starts a chain that builds new connections. Two proteins do the work. The first is called BDNF, and it acts like a fertilizer. Where there is more of it, neurons grow and their connections thicken. The second is called mTOR, and it acts like a builder. It tells the cell to make the materials and put the new connections in place. Psychedelics turn on both. When researchers grow neurons in the lab and shut off either protein, no new connections form, even with the drug present. That is how researchers know the growth runs through these two proteins, and not by some other route. The growth runs in a short burst, a few hours, then stops. That short burst is enough. Once the new wiring forms and gets used, it can last long after the drug and the growth signal are gone.
All of that is the repair side: the inside receptors, BDNF and mTOR, the new connections. The open question is whether the trip runs on this same machinery or on a path of its own. If the trip and the repair really do run on separate machinery, it should be possible to engineer a molecule that does one and skips the other. The first convincing proof came from an unpromising raw material, ibogaine, a psychedelic drawn from the root bark of the iboga shrub of West Africa. Ibogaine has a long folk reputation for breaking heroin and other addictions in a single brutal session. But it is genuinely dangerous. It disturbs the heart’s rhythm in ways that have killed people. Olson’s lab took the molecule apart and rebuilt it as a simpler synthetic version, removing the structural features tied to its toxicity. They named it tabernanthalog.
In rodents, tabernanthalog did the good things and skipped the bad one. It grew new branches and connections in cortical neurons, the structural sign of repair. In a standard test of despair, a single dose made the animals keep struggling rather than give up, the signal of an antidepressant effect. In addiction models it cut their drive to seek alcohol and heroin, and the benefit outlasted the drug’s brief presence in the body. Later work showed a single dose could also reverse the anxiety and rigid thinking that chronic stress stamps into a mouse, and restore the circuits stress had disrupted. It proved far gentler on heart cells than ibogaine. And it produced no head-twitch. By the field’s own ruler for the trip, the molecule registered nothing. That was the first solid evidence that the two effects can be pulled apart, that one compound can drive the plasticity while staying silent on the hallucination, at least in a mouse.
Everything hangs on that last phrase, at least in a mouse. The structural hypothesis is elegant, internally consistent, and supported by a body of animal work that grows more detailed every year. But a mouse cannot say whether it feels better. It can only swim a little longer in a tank before it quits, or reach for the drug a few times less. The rest is inferred. The leap from a rodent’s behavior to a depressed person’s recovery is exactly where the confidence gets thin. It is still genuinely unsettled whether a human brain needs the conscious trip itself, the hours of dissolved perception and raw emotion, to turn fresh wiring into lasting relief. Or whether the wiring is enough by itself, and the trip is just noise that can be deleted. The animal data show the two can be separated in principle. Whether they come apart in a person, and whether the silent version still heals, is the question everything else depends on.
The Window, Not the Cure
A child’s brain can do things an adult’s brain cannot, and the gap has nothing to do with effort. Early in life there are windows when certain circuits are wide open to being shaped by experience. They take their final form based on what comes in, and then they close. Once they close, they stay closed for good. Language is the familiar example. A young child exposed to any language on earth absorbs it whole, grammar and accent, without seeming to try. An adult studying that same language drills for years and still speaks it like a foreigner. The machinery for picking up a first language runs on a clock, and the clock runs out.
The clearest proof came from cats. In the 1960s, two neurophysiologists at Harvard, David Hubel and Torsten Wiesel, sewed shut one eye of newborn kittens for a few weeks early in life. The two would later share the 1981 Nobel Prize for mapping how the visual brain is put together. The part of the brain that normally blends the two eyes into a single three-dimensional image rewired itself to ignore the shut eye and serve only the open one. When they reopened the closed eye, it was perfectly healthy and its signal reached the brain intact. But the territory was already lost. The cat was permanently blind in that eye. The same procedure on a grown cat changed nothing, because the adult wiring was already set. In humans the same window is why a child’s “lazy eye” has to be corrected early; once it closes, the vision in that eye is, for practical purposes, lost.
Why would a window close and refuse to reopen? Part of the answer is that the brain physically clamps its own wiring down. As a circuit finishes learning, certain neurons get wrapped in a dense lattice of molecules called a perineuronal net. Think of it as a cage that builds up in the space around the cell. While that space is still loose, connections can be added, pruned, and rearranged. Once the net hardens, the layout is locked in place, the way wet concrete can be shaped for a while but holds its form for good once it sets. So the net is a brake on change. It is the brake itself, rather than a symptom of aging. When researchers chemically dissolve these nets in an adult animal, the long-closed window opens back up and the circuit can be rewired again. A critical period, then, is not lost because neurons die or wear out. It is held shut by a structure. And a structure can be removed.
For decades, reopening one of these windows looked like it would require exactly that kind of brute intervention: dissolving the cage by hand in one small patch of brain. Then a neuroscientist found that a single dose of a familiar drug could open the lock from the inside.
Gül Dölen was an associate professor of neuroscience at Johns Hopkins, and in January 2024 she moved to an endowed chair in psychedelic science at the University of California, Berkeley. She had spent years on one particular window, which she calls the social reward learning critical period. Young mammals, mice and humans alike, go through a phase when they are most open to social bonding. The company of others feels most rewarding, and the lessons about belonging cut the deepest. In humans this phase is adolescence. In mice the window is sharp and easy to test. A juvenile mouse quickly learns to prefer a place where it once had the company of other mice. But an adult mouse, run through the same test, has lost the ability. The window has shut.
In 2023 Dölen’s lab reported that MDMA reopened it. A single dose gave an adult mouse back, weeks later, the juvenile ability to learn that being around others is rewarding. This is a capacity that developmental neuroscience had long believed was gone for good once the window closed.
Then came the result that changed the whole field. The effect was not special to MDMA. Psilocybin reopened the window. So did LSD. So did ketamine and ibogaine. These two share neither the chemistry nor a serotonin receptor with the others, and they produce nothing like the same kind of trip. Five unrelated drugs, and every one of them pried open the same lost window of learning. Dölen’s reading was that this, and not any single chemical pathway, is what psychedelics really do. They are a master key to critical periods. What they restore is not one fixed circuit, but the brain’s general readiness to be rewired by experience.
But the most interesting part came next. The window did not stay open for the same length of time with every drug. After a single dose, the social learning window stayed open in mice for roughly 48 hours with ketamine, about two weeks with psilocybin and with MDMA, three weeks with LSD, and four weeks with ibogaine. The drugs lined up on a scale running from two days to a month.
And the ordering was not random. It tracked, roughly, how long each drug’s acute effects last in a person who takes it. Ketamine clears fastest. An ibogaine experience can run a full day or more. The rest fall in between, in the same order as the windows they open.
The implication is striking. The length of the open window may be something you can choose, rather than a fixed trait of the drug class. A treatment that needs only a brief, intense burst of relearning might call for a short-acting drug. One that needs weeks of practice and consolidation might call for a long-acting one. You could choose the window length to fit the task, the way a cast stays on for exactly as long as the bone needs and no longer.
Seen side by side, the window picture and the wiring-repair picture change the role of the drug. The drug is not the cure, but what opens the window. The cure is whatever you learn, practice, or work through while the window is open. A pried-open critical period is an opportunity. It is a span of days or weeks during which the brain will take in new experience and lock it into the wiring. If the right input comes during that span, the brain wires in the repair. If nothing comes, the opportunity is wasted.
This one shift rescues something the older, mystical accounts got right for the wrong reasons. Therapists who work with these drugs have always insisted on “set and setting,” the patient’s state of mind and the physical and emotional surroundings of the session: the music, the safety, the guide, the intention. The mystical reading treated this as respect for a sacred experience. The window reading turns it into mechanism. If the brain is at its most teachable for a stretch of time, then the experience you deliver during that stretch is the active ingredient. A frightening or chaotic setting is then not a failure of respect, but a mistake in the treatment itself.
Dölen draws the blunt conclusion: the right input depends on the disorder. A guided introspective journey is the correct content for trauma, addiction, or depression, where what has to be relearned is emotional and relational. It is the wrong content for stroke. There the open window should be filled with physical rehabilitation, the patient drilling the lost movement while the motor circuits can briefly rewire. The drug opens the window. The condition decides what to send through it.
This reframe cuts in two directions at once. It raises the stakes for the whole approach, because the drug by itself is only half a treatment, even a clean, non-hallucinating one. Suppose a patient takes a perfect non-tripping drug that opens a plasticity window, then goes home to a stressful apartment and a steady stream of bad news. The open window may faithfully record the stress. A brain pried open to learning is also a brain pried open to learning the wrong things.
Plasticity has no preference for improvement. It records whatever the environment supplies. The same openness that lets a trauma patient relearn safety could, in a hostile setting, cut the injury deeper. The window is a powerful tool, not a guaranteed cure. An open window left unattended can do harm as easily as good.
This is where the trip comes back into the argument, the same trip the no-trip camp wanted to remove. If an open window has to be filled with structured, carefully shaped experience to do any good, then the vivid, emotional, introspective hours a psychedelic produces may not be a side effect to engineer away at all. They may be the structured input itself: the very experience the brain absorbs and locks into its wiring during the brief time the window stays open. A non-hallucinating drug would open the same window but arrive with nothing to put in it, leaving whatever the brain builds to whatever the patient happens to encounter.
That leaves the field with one question it cannot yet answer. Once a brain has been pried open, can it be trusted to fill the window on its own, or does it need a guide to shape what goes in? The window itself is real; what no one can promise yet is that prying it open, by itself, is either safe or enough.
Three Ways Past the Session
Suppose the chemistry is solved. Suppose a molecule reliably regrows the brain’s lost wiring in people, with the hallucination set to whatever length a company wants. There is still a wall between that molecule and a mass market, and it has nothing to do with chemistry. The wall is how the drug has to be delivered.
Look at how the delivery works today. A patient swallows a dose of synthetic psilocybin. Then they need a quiet, controlled room, usually eyeshades and music, and one or two trained monitors sitting beside them for six to eight hours while the visions run their course. The session cannot be rushed, because the drug sets the clock. In Oregon, where supervised psilocybin became legal outside the federal system, a course of treatment costs roughly $2,000 to $3,000. In Colorado it costs $4,000 to $5,000. The most therapy-heavy protocols, the ones that wrap many hours of preparation and follow-up talk around the dose, have been priced as high as roughly $9,500. These prices come from labor. A single patient can take up about ten hours of a trained facilitator’s time, across preparation, the dosing day, and the talking-through afterward, and the dosing day itself often needs two guides in the room at once. None of this is paid for by insurance today.
That is the bottleneck. It is operational and financial, not chemical. Trained monitors are scarce, and the training is neither quick nor cheap. When Australia became the first country to let psychiatrists prescribe psilocybin and MDMA, the math of scaling became brutal. Treating even a thousand patients in a year would need more than a hundred authorized psychiatrists and over a thousand therapists. By the end of 2025 the country had cleared only about thirty psychiatrists for the job. A health system built around the fifteen-minute appointment and the bottle of daily pills has no place for a supervised eight-hour session, and no settled way to pay for one. A drug that works perfectly and still needs that session is a drug that reaches thousands, not millions.
This is the constraint every serious company in the field is now racing to solve, and the molecule was never the hard part. Three camps have formed. The cleanest way to tell them apart is by what each one does to the session. One keeps it. One shortens it. One tries to delete it.
The company keeping it is the one winning the race. Compass Pathways carried synthetic psilocybin through two large successful trials to the front of the regulatory line, which means it will most likely be first to sell a classic psychedelic as approved medicine. But the win comes with a burden. COMP360 still produces a session of six hours or more that has to be supervised, and the commercial job now is to build the system that delivers it at scale and gets paid for it. Compass’s own plan leans on roughly seven thousand existing interventional psychiatry centers, the clinics already equipped to give multi-hour treatments, as the rooms its sessions can fit into. There are gates to clear that have nothing to do with whether the drug works. Regulators can attach what is called a risk evaluation and mitigation strategy, a safety program that restricts who may prescribe and dispense a drug and requires the monitoring around it. The Drug Enforcement Administration would have to move psilocybin off the most restrictive tier of banned substances before any doctor could write a prescription. The working template for all of this is the esketamine nasal spray already stocked in those clinics. Compass has to ride that template while carrying the longest session of anyone in the field.
The second camp accepts the supervised session but goes after its length. The molecule’s grip on the brain is what sets the clock. So a molecule that grips briefly turns eight hours in a clinic room into something closer to an ordinary appointment. The compound of choice is 5-MeO-DMT, a fast and overwhelming psychedelic whose full effect arrives and fades inside fifteen to thirty minutes rather than across most of a day. GH Research, a biotech listed on Nasdaq, delivers it as a vapor the patient inhales, under the laboratory name GH001. In its mid-stage trial in treatment-resistant depression, a short course of these brief sessions drove a large drop in symptoms, and the number that mattered was durability. Roughly three-quarters of patients were still in remission six months later, after an average of only four treatments, and with no structured psychotherapy wrapped around the dose. The Food and Drug Administration had paused the program over a safety question, then lifted that hold, clearing a global final-stage trial to begin in the second half of 2026. The pitch is the same relief delivered in a small fraction of the chair time.
A near-twin bet sits right beside it. atai Beckley, formed in November 2025 when atai Life Sciences absorbed the British developer Beckley Psytech after seeing its trial data, is running a nasal spray version of the same 5-MeO-DMT molecule under the name BPL-003. In its mid-stage depression trial a single dose produced rapid relief that held for weeks, and most patients were ready to walk out of the clinic about ninety minutes after dosing. In October 2025 American regulators granted it Breakthrough Therapy status, a designation that flags a drug as a possible major advance over what exists and pulls extra agency attention to hurry it along. The aim is to fit the whole treatment inside the same interventional-psychiatry footprint the esketamine spray already occupies, the ninety-minute visit in place of the eight-hour session. This is the pragmatic middle of the field. It does not argue about whether the trip heals. It shrinks the trip until the session stops being the thing that breaks the economics.
The third camp makes the most radical claim: remove the trip and the session vanishes with it. Delix Therapeutics, still privately held and founded by David Olson, leads it. Its lead molecule, zalsupindole, is built on the same 5-MeO-DMT skeleton that GH Research vaporizes and atai Beckley sprays, but reworked into a pill the gut absorbs and stripped of the structural features that produce hallucination, sedation, and dissociation. It is a non-hallucinogenic neuroplastogen, engineered to drive the rewiring while leaving the patient’s perception untouched. In October 2025 Delix reported that an early human trial showed both the chemical fingerprints of fresh plasticity and rapid, lasting relief of depression. What mattered more than the results was what the agency then allowed: a mid-stage trial in which patients dose themselves at home, one arm taking a daily pill and another taking one twice a week. No room, no monitor, no clinic. It is the most complete answer to the bottleneck anyone has given, and also the one furthest from market, an early-stage signal rather than a finished drug. The boldest bet in the field is also the least proven.
Laid out as three sharp camps, the field looks tidier than it is. In practice it is a spectrum, and the gaps between the camps are filling with companies that tune the experience rather than keep, shorten, or delete it. Two are well into final-stage testing. The first spent years as Cybin and took the name Helus Pharma when it moved to Nasdaq at the start of 2026. It is developing a modified form of psilocin, the active compound the body produces from psilocybin. The modification is deuteration. Deuterium is a heavier, sturdier twin of ordinary hydrogen. Swapping it onto the molecule at the right spot makes the body break the drug down more slowly, so a smaller dose reaches the brain more efficiently and behaves more predictably. The trip itself remains, in something close to its full form, with a late-stage depression readout due late in 2026.
The second of the tuners spent years as MindMed and is now Definium Therapeutics. It is developing LSD, the molecule it all began with, formulated into a tablet that dissolves on the tongue. Three final-stage trial readouts are due across 2026, led by generalized anxiety and reaching into depression. Neither of these companies is trying to delete the experience. They are polishing the oldest molecules in the field, betting that a cleaner, more predictable version of a full trip holds its own market even while rivals race to shrink the trip away. Place every company on one line and it runs without a gap from a full six-hour psilocybin session at one end to a pill swallowed at home with no supervision at the other.
The commercial logic underneath that line is simple. A treatment that needs a clinic, a monitor, and most of a day is expensive to deliver and hard to get paid for. A treatment a patient takes at home collapses both problems at once. The cost of the session falls toward zero, and the drug starts to look like any other prescription an insurer already knows how to reimburse. The further down the line a company sits, toward no session, the larger the population it can reach, and the more its product looks like something a health system can actually absorb. So the question of who shortens or removes the session is really the question of where the money lands. And the order of the camps on that line is close to the order of the markets they are reaching for.
That points to a layer of value that gets less attention than the molecules do. The molecule is the part everyone writes about. The durable asset may instead be the delivery system wrapped around it: the certified clinics, the monitoring protocols, the reimbursement codes, the at-home safety regime, the machinery that decides whether an approved drug reaches thousands or millions. And the clearest read on where experienced money thinks the value sits comes from watching who is buying. A decade and a half ago, several of the largest drugmakers shut down their brain and psychiatry research, judging it too slow and too uncertain to fund. Psychedelics are one of the doors some are now using to come back, and the cheapest way back in is to buy a small company that has already done the hard clinical work. AbbVie, a large drugmaker with an established psychiatry business, has placed its money at the shortened and deleted ends of this line rather than the keep-the-trip end. The buyer with the most to spend is reaching for the compounds that shrink or delete the session, not the one that is first to approval with the session left intact.
None of this settles who wins, and it is not meant to. The front-runner could prove that a supervised psilocybin session, once it is properly reimbursed, scales further than the skeptics expect. The take-home pill could stall in its next trial, or work in the lab and then quietly rewire the wrong things in an unsupervised living room. What this shows is where the bets sit and why. It locates the one constraint every camp is built around, the session. And it shows that the value flows to whoever can loosen that constraint without losing the repair. Where on the line it lands, and which name holds the right asset when a larger buyer comes calling, is the wager. The reader now has enough to weigh it.
The Gatekeepers and the Back Door
The science can show that a molecule regrows the wiring in a brain. But it can’t answer a different question. Can the institutions that license drugs, pay for them, and hand them to patients find a place to put one? That answer gets decided not in a laboratory but inside regulatory agencies, insurance companies, and state legislatures. So far the answer is no. The system does not fit these drugs, and it’s now being forced to bend around them, in ways that are fast, political, and far from settled.
Start with why the fit is bad. The system that approves a drug in the United States was built to do one thing: isolate the effect of a single molecule and measure it, with everything else held constant. The tool for that is the blinded trial. You split patients into two groups. One gets the drug. The other gets an inert look-alike, the placebo. Neither the patients nor the doctors scoring them know who got which. Hope, expectation, and the simple belief that you are being treated are strong enough to lift symptoms on their own. So the blind exists to strip them out. Whatever benefit is left in the drug group, after you subtract the placebo group’s improvement, is taken to be the molecule’s true effect.
A psychedelic breaks that design. Within an hour of dosing, the patient knows exactly which group they are in, because the walls are breathing. The blind collapses, and expectation floods back into the result the blind was built to remove. A patient who knows they got the real drug, and who often signed up for the trial exactly because they were hoping for it, will tend to report feeling better whether or not the molecule did anything. There is no clean way to subtract that out.
The second problem runs deeper. What gets tested is not a molecule. It’s a molecule plus many hours of psychotherapy: the guided preparation, the supervised dosing session, and the talk afterward. The agency approves drugs. It doesn’t license psychotherapy, which is overseen by state boards and professional bodies and varies from one therapist to the next. A treatment that is half pill and half relationship has no clean home in a system built to evaluate the pill. Approve the molecule alone, and you have approved something that was never tested by itself.
Lykos Therapeutics showed what this bad fit does to a real application, in public and at length. Lykos is the corporate arm spun out of the nonprofit that had spent decades pushing MDMA as a treatment for post-traumatic stress disorder. It carried the first such application all the way to the finish line, and it was taken apart there. In June 2024, the agency’s outside panel of experts, the Psychopharmacologic Drugs Advisory Committee, voted nine to two that the company’s data did not establish that the drug worked. It voted ten to one that the benefits did not outweigh the risks.
The reasons trace right back to the structural problem. Functional unblinding was not a worry in the abstract here. It was measured. In the main trial, more than ninety percent of the patients who got MDMA correctly guessed that they had. So did about three-quarters of those given the placebo. Almost everyone knew. On top of that, the trials drew heavily from people who were already enthusiastic about the drug. Around forty percent of participants had taken MDMA before. And anyone whose earlier experience was frightening was unlikely to volunteer. That loaded the studies with patients who leaned toward a good result. Then came conduct problems that had nothing to do with chemistry. One participant in an earlier study reported sexual misconduct by the therapist administering the treatment and her unlicensed husband. A drug-pricing watchdog alleged that some therapists in the program had steered patients toward reporting good outcomes and away from reporting the harms they suffered. In a treatment where the therapist is part of the dose, the therapist’s conduct becomes trial data. And there was no established way to police it.
The verdict arrived on August 9, 2024, as a Complete Response Letter. That’s the document the agency sends to say no. It declined to approve the drug and told Lykos to run another full late-stage trial, a fresh Phase 3, before trying again. The letter also faulted the psychotherapy wrapped around the drug. It was too loosely defined to be reproduced reliably if approved. This is the drug-versus-therapy problem, stated in the agency’s own words. The company’s response was not a minor adjustment. Within a week it cut about three-quarters of its roughly one hundred employees. Rick Doblin, who had founded the parent nonprofit in 1986 and pushed the cause for nearly four decades, stepped off the board. That is what the standard process does to a psychedelic. The trip makes a clean trial nearly impossible, and the therapy is half the treatment. It destroys the application.
And yet the same agency, less than two years later, is bending hard the other way for a different psychedelic. The structural problem did not go away. Psilocybin announces itself in a patient just as plainly as MDMA does. Its trials carry the same broken blind. What changed was the institutional stance toward that evidence. Compass Pathways is being allowed to file its psilocybin application in pieces and carry it through on the agency’s fastest track. And the new friendliness extends well past one company.
On April 18, 2026, a presidential executive order titled Accelerating Medical Treatments for Serious Mental Illness directed the FDA to give psychedelics expedited review. It also ordered the federal health department to put fifty million dollars toward matching state spending on psychedelic research. The agency then issued three of its National Priority Vouchers to psychedelic developers at once. Compass took one, for psilocybin in treatment-resistant depression. The nonprofit Usona Institute took a second, for psilocybin in major depressive disorder. Transcend Therapeutics took the third, for methylone, a chemical cousin of MDMA, in PTSD. Two psilocybin indications and one PTSD drug, all moved to the front of the line.
Put the two moments side by side. The same regulator that forced Lykos to cut three-quarters of its staff in 2024 was, by 2026, clearing a path for three psychedelic programs at once. And the impossible blind that sank the first was no more solved than before. The science had not improved, but the stance had. And a stance is a variable, not a constant. It can swing on an election, a new commissioner, a single executive order. This means the agency’s current friendliness is a fact about this political moment, not a fixed foundation a company can build a decade on. Support that arrived this fast can disappear just as fast.
Approval, when it comes, settles less than it looks. A drug that clears the agency has won the right to be sold, not the right to be paid for. And for a supervised psychedelic, the bill is not mostly the molecule. It’s the molecule plus the room, plus the chair the patient sits in for most of a day, plus the trained people who have to sit there too, plus the hours. Someone has to agree to cover that whole stack. And the party that decides is the insurer, not the agency.
The closest thing to a working model is the esketamine nasal spray already used for hard-to-treat depression. It can only be given inside a clinic certified under a restricted safety program, with the patient watched for about two hours afterward. The way those clinics get paid is the template every psychedelic company is studying. For patients on the government’s older-adult insurance, the clinic bills a single bundled code. It folds the drug, the supervision, and the two-hour observation into one payment of roughly thirteen hundred dollars. Commercial insurers instead pay for the drug by the milligram and add a separate charge for administering it. A session runs somewhere between seven hundred and twelve hundred dollars. The government plan covers most of it once the patient has met the annual deductible.
The instructive part is how long all that took to build. The spray was approved in 2019 and barely sold for years. Its maker did not even break out the revenue at first. The billing codes, the payer contracts, and the network of certified clinics took until roughly 2023 and 2024 to mature. Only then did the product turn into a real blockbuster. Sales reached about $1.1 billion in 2024, the spray’s first year past that mark, then climbed roughly 57 percent to nearly $1.7 billion in 2025, while the number of certified sites grew from about 2,800 to more than 7,000 in two years. Roughly five years passed between approval and real scale. And that was for a two-hour session that already had its codes written. A psilocybin session runs three times as long. And no billing code for a six-hour supervised drug session exists yet. One would have to be created from scratch. Approval does not end the race. It starts a second, slower one, run inside insurance companies, over whether anyone will pay for the supervised session.
While the medical track inches through approval and billing codes, a second track has opened. It routes around the agency entirely. Voters in two states have legalized supervised psychedelic use, with no doctor, no diagnosis, and no prescription. Oregon opened the first such service centers in the summer of 2023. By 2026, something like sixteen thousand people had taken psilocybin inside them, watched by a licensed facilitator rather than a physician. Colorado followed, under a law it calls the Natural Medicine framework. It held its first state-regulated session in Denver on June 6, 2025. That was a five-hour high-dose experience, with required preparation and counseling on either side. Its advisory board is now weighing whether to widen the menu to three more substances: ibogaine, DMT, and mescaline, the psychedelic drawn from certain cacti.
This is not medicine, and it is not trying to be. There is no insurance code because there is no diagnosis. There is only an adult who wants the experience and pays cash for it. The model could capture exactly the demand the pharmaceutical track is too slow or too expensive to serve. These are the people who will not wait years for an approval, or who fall outside the narrow condition a drug gets licensed for. But it inherits the same cost stack, with none of the insurance to absorb it. And the strain shows. Oregon’s centers charge thousands of dollars out of pocket, and many have not survived it. By 2025, of the roughly thirty-five centers that had been licensed, only about two-thirds were still operating. The rest closed under the weight of high costs and heavy regulation. The route around medicine is real, but it is financially fragile.
What all of this leaves is a genuinely forked future. And the science does not get a vote on which fork wins. Down one path, these drugs become ordinary medicine. They’re approved by the agency, prescribed by doctors, delivered in certified clinics under a safety program, and eventually reimbursed by insurance. That path is expensive, slow, and supervised. But it carries the legitimacy and the reach that come with running inside the health system. Down the other path, the same drugs are reached through the state-regulated or decriminalized track. This is outside medicine, paid for in cash. It’s faster to enter and broader in who it admits, but unreimbursed and legally patchwork, different in every state.
The most likely outcome is not that one path wins. It’s that both exist at once and barely touch. A patient with a diagnosis of treatment-resistant depression and an insurance card flows toward the clinic and the prescription. A healthy adult who wants something the medical system would never write a code for, or a patient priced out of that system, walks into a service center instead. The two tracks would serve different people, run on different money, and rarely cross. That split matters more than any single trial result for where the impact and the value actually land. The science can show that one of these molecules rewires a brain. It cannot say who will be allowed to sell it, to whom, or who will agree to pay. Those are institutional questions. And they are the ones still genuinely open.
More Than Psychiatry
Depression is only the wedge. It is the first door because that’s where the evidence is thickest and the trials are largest. But if the deepest version of the plasticity idea is right, that these molecules are general-purpose tools for reopening the brain’s ability to be rewired, then mood is only the first thing they touch. The same machinery would reach learning, recovery from injury, addiction, and the slow decline of the aging brain. The temptation is to say all of this in one breath, as if proof of the wedge were proof of everything beyond it. It isn’t. Each step away from depression is also a step down in certainty. The honest way to take these steps is to climb deliberately, and to name each rung for what it is: firm, early, a leap, or a guess. What follows climbs from the most solid footing to the thinnest.
The first rung is the firmest, because it is less a prediction than a way of organizing what is already known. The future it describes is not one drug but a menu. The same window-duration dial, short-acting compounds for a brief burst of relearning and long-acting ones for weeks of practice, is itself the platform vision. Under that picture, psychiatry stops looking like a daily pill that holds symptoms down for as long as the patient keeps taking it, and starts looking like a scheduled procedure: open the window, deliver the targeted therapy or training or rehabilitation, let the window close, and leave the repair in place. It is more like a course of physical therapy, or a surgery with a set recovery time, than a drug you take forever. That is the platform vision, and it rests on evidence in hand, the window data and the wiring data both. It asks only that the pieces be put together. Plausible.
The next rung still stands on real human trials, but only their earliest ones. Addiction is the natural first target, because it looks less like damaged wiring than like a rut. Once the window reopens, the rut can be rearranged. In 2022 a team at New York University led by the psychiatrist Michael Bogenschutz ran a controlled trial in people with alcohol use disorder. Two psilocybin sessions, measured against a placebo, cut the share of heavy-drinking days by more than half in the following months. The smoking signal is older, and given the small numbers, even more striking. In a small Johns Hopkins study, a few high doses taken with counseling left about 80 percent of entrenched smokers abstinent six months later, a number the lead researcher called surprising. These studies are small, and the broken blind that troubles every psychedelic trial troubles these too. But the benefit lands where the theory says it should, a few doses loosening a fixed circuit that daily medicine barely moves.
The most literal use of an open window is not psychiatric, but repair after physical injury. When a stroke kills a patch of motor tissue, there is a brief natural window in which the surrounding brain can rewire to take over the lost function. It closes within weeks. A drug that reopens it could make an injury years old treatable again. In June 2025 a Johns Hopkins team that included Dölen and the stroke-recovery neurologist John Krakauer opened the first trial of exactly that, pairing a dose of psilocybin with intensive, digitally guided motor training in patients left with chronic disability. Here what fills the open window is physical practice, not introspection. That makes the logic of the whole platform plain: the drug opens the window, and the rehabilitation is the cure. It is a Phase 1 study, the first rung of testing in humans, built to check safety before it can say anything about recovery. Related work is moving toward spinal-cord injury. Early, but real.
The boldest extension is not a new disease to treat, but a new way into the cell. The growth a psychedelic produces is carried by BDNF, the protein that acts as fertilizer for neurons. BDNF works by fitting into a receptor on the neuron, a lock named TrkB, and turning that lock flips the switch that tells the cell to grow. The standard account runs the whole effect through the serotonin 2A lock, which then, several steps later, produces BDNF. A competing account argues that the drugs skip most of that detour and go straight for the growth switch.
It comes from the lab of Eero Castrén, a neuroscientist at the University of Helsinki who has spent his career studying how antidepressants reopen plasticity in the adult brain. In 2023 his group reported that LSD and psilocin bind TrkB directly, and grip it about a thousand times more tightly than ordinary antidepressants do. In their mice, the plasticity and the antidepressant effect depended on TrkB binding and did not need the serotonin 2A lock. The head-twitch that marks the trip was the reverse: it needed serotonin 2A, not TrkB. Two years earlier the same group had shown something stranger still: that the everyday Prozac-family antidepressants, and ketamine, also work in part by binding TrkB. If that is right, the growth switch can be flipped directly, without first turning the lock that produces the hallucination. The old drugs were running a slow, gentle version of the same trick all along. A molecule built to hit only TrkB would be a plastogen with no trip to engineer away, because none was ever turned on. The account is contested, and the two leading labs disagree about which lock matters most. But it points to a second door into the same room, more direct than the first.
Higher up the ladder, the footing gets weak. This is the leap, and it should be called one before anything else. The target is not the wellness industry’s “longevity” but two concrete features of the aging brain: the loss of plasticity, and neuroinflammation. The brain keeps its own immune cells. When they run hot for years, a low fire that never goes out, that chronic inflammation is among the suspected causes of Alzheimer’s. The drugs appear to act on both fronts. Charles Nichols, a pharmacologist at Louisiana State University, found that psychedelics are powerful anti-inflammatory agents, and that this effect shows up at doses far below the trip, through the serotonin 2A lock turned in a particular way. So a single class of molecule might both reopen plasticity and cool the fire, the two things a slowly failing brain most lacks. A 2025 study in an Alzheimer’s mouse model found that psilocybin lowered brain inflammation and drove the growth of new neurons in the hippocampus, the brain’s memory center. A case report published in 2026 described an 80-year-old woman whose Alzheimer’s had narrowed her speech to single syllables for years. After a heavy dose of psilocybin mushrooms, she spoke in sentences again, briefly.
That is exactly where discipline matters most, because a mouse and a single patient are not evidence that anything works. They are reasons to run a trial. The human record in older patients is close to empty: across roughly 1,400 people enrolled in modern psychedelic trials, only about 19 were 65 or older. And the old are the group most exposed to the drugs’ real risks: the jump in blood pressure and heart rate that strains a fragile cardiovascular system, the danger of a fall during hours of altered perception, the interactions with all the other medicines an older patient is likely taking. The cognitive-healthspan idea is the most exciting rung and the least supported by anything close to efficacy data. It is a leap, and calling it a leap is the price of taking it seriously.
The highest rung is not a medical claim, but a cultural one. Part of it is simply removing the stigma, the slow turning of a banned street drug into something a doctor prescribes and an insurer pays for. The bigger part is a shift in what mental healthcare is for. Today it waits for a person to break, then treats the break. A safe, occasional, take-home plastogen, the kind the at-home dosing trials are moving toward, would invite a different habit: maintenance before the break, a kind of mental fitness done on a schedule, the way people exercise to stay healthy. This is a social projection, not a finding about any molecule, and it should be named as one. It could arrive even if the drugs turn out to do less than their backers promise. It could fail to arrive even if the drugs work, because access, cost, law, and culture decide it, not pharmacology. It belongs on the ladder, but as a different kind of claim, the kind no trial can ever settle.
The cleanest test of whether the field can hold its line runs through the part of it that has already run ahead of the evidence. Microdosing is taking doses too small to produce any noticeable effect, on a regular schedule, to get sharper focus, steadier mood, more creativity. The popular enthusiasm is enormous. The controlled data point the other way. In the largest placebo-controlled study, 191 people blinded themselves by mixing real and dummy capsules they could not tell apart. Both groups improved, and neither beat the other. A separate double-blind trial that handed volunteers half a gram of dried mushrooms or a placebo found no gain in mood, thinking, or creativity, and if anything a few small changes for the worse. Expectation, not the molecule, explained the lift people reported feeling. The contrast is worth holding onto. The doses that regrow wiring in the clinical trials are large, infrequent, and supervised. The wellness microdose is small, daily, and self-administered, and it does not survive a placebo control. Saying that plainly is what keeps every rung below it credible.
Stand back, and the structure is clear from bottom to top. The platform is plausible, built from evidence already in hand. The reach beyond mood, into addiction and physical injury, is real but still in its first human trials. The healthspan promise is a real leap, balanced on a mouse, a case report, and a population almost no study has touched. The cultural shift is a projection no experiment will decide. And the wellness claims at the foot of the ladder collapse the moment you add a placebo. Depression is the wedge because depression is where the proof is. Everything stacked above it is a bet that one trick, reopening the adult brain to change, reaches further than a single diagnosis. The bet may well pay off. What the field owes, and what separates a science from a faith, is to keep saying out loud how high up the ladder each claim sits.
The Cracks in the Thesis
Charlie Munger, the investor, spent his whole career preaching one idea. The way to handle a hard problem is to invert it, to study how a thing fails before asking how it might succeed. He left a useful instruction for a moment like this. Lay out every way the structure could collapse, make each failure as strong as its sharpest critic could make it, and only then count what’s left standing. The plasticity thesis is a real possibility. But the discipline that earns you the right to believe any of it is the willingness to honestly take it apart first.
Start with the deepest threat, the one aimed at the foundation of the no-trip camp. That camp’s entire bet is that the conscious experience is engine noise, a side effect you can delete while keeping the repair. The evidence runs the other way, and it’s uncomfortable. Every molecule that has produced a clean antidepressant signal in an actual human so far is a molecule that makes that human trip. The one closest to becoming medicine kept the trip in full. Worse for the skeptics, the psilocybin depression trials point the same way. The patients who reported the most complete and overwhelming experiences, the full dissolution of the ordinary self, tended to be the ones who got the most lasting relief. This correlation has shown up in more than one study. Now, this doesn’t prove the trip is the medicine. The intensity of the experience might be nothing more than a readout of the dose, or of how thoroughly the drug engaged the brain. But there’s also a clean theoretical reason it might matter. A brain briefly thrown open to relearning still needs something to learn, and the vivid hours of a trip are the most structured input these drugs deliver. The case that the experience is optional rests, for now, on rodents. The case that it matters has human beings behind it.
This points straight at the second crack, which runs underneath the whole decoupling project. To claim that a molecule heals without tripping, you have to measure both halves in an animal first, because the earliest trials can’t be run any other way. So the field leans on two substitutes. The side-to-side flick of a mouse’s head is used in place of human hallucination. The count of fresh connections on a rodent’s neurons is used in place of a depressed person’s recovery. Proxies like these have failed many times before. The clearest warning comes from stroke. Over the past few decades, more than a thousand candidate treatments protected brain tissue in animal models of stroke, and essentially every one of them then failed in humans. One of the most promising was a free-radical scavenger called NXY-059. It passed its rat and primate studies and an encouraging first human trial, and then a second, larger trial erased the result and the drug was abandoned. The reason was mundane and damning. A rat’s stroke sits in different tissue than a human’s, and the model had been making the drug look better than it was the whole time. Fewer than one in ten brain drugs that enter human testing ever reaches a pharmacy. A mouse’s twitching head looks like a human trip. Whether the absence of that twitch predicts a human who heals without one is a bet the animal data cannot settle.
Suppose the molecules clear that bar and work in people. A third question waits. Does the new wiring hold, or does it loosen and let the patient slide back? The honest answer from the strongest data is mixed. A single dose of psilocybin can lift depression for months. But in the larger controlled trials, the benefit that is clear at three weeks is no longer reliably there at twelve, and across a full year the effect decays in a meaningful share of patients. Two doses hold better than one. So this may not be the one-and-done cure the early excitement imagined. It may be a treatment that has to be repeated, a course rather than a single shot. That matters for the business, because a therapy taken once is a far smaller market than a therapy taken again and again. And it matters for safety, because repetition is exactly where a quieter danger begins.
This danger sits on a different receptor. Brain cells are not the only place the body keeps serotonin receptors. The valves of the heart carry their own variant, labeled 5-HT2B. When it is held open for long stretches, the valve tissue thickens, stiffens, and stops sealing the way it should. This is not hypothetical. It is the exact mechanism that pulled the diet-drug combination fen-phen off the American market in 1997, after it scarred the hearts of thousands of users. The same flaw later forced the withdrawal or tight restriction of several Parkinson’s and migraine drugs. Classic psychedelics grip this same receptor. Taken once or a handful of times, with long gaps in between, the exposure is probably too brief to do harm. A century of intermittent use has produced no epidemic of damaged valves. But a drug taken to hold off relapse, dosed every week or, in the wellness fantasy, every few days, is a different kind of exposure, and no one has run the study that would clear it. Add to that the sharp temporary spikes in blood pressure and heart rate that every dose produces. Then there’s the absolute bar against anyone with a personal or family history of psychosis, in whom these drugs can trigger a psychosis that does not lift. The eligible population narrows while the liability climbs.
Even a molecule that proved safe and durable would hit an obstacle that has nothing to do with biology. The compounds at the center of all this are old. LSD was first made in 1938, psilocybin was synthesized and characterized in the late 1950s, and the tryptamines are older still. No one can own a molecule the world has known for three generations, which is why almost no one is trying. The protectable asset is everything wrapped around the molecule: a chemical tweak that makes it last longer or absorb as a pill, a particular inhaled or dissolved formulation, a specific dosing schedule, the certified clinics and billing codes and safety programs. These are real forms of advantage, but they are thinner and more contestable than a patent on a brand-new compound, and a determined generic competitor can sometimes engineer around them. A field built on public-domain chemistry has to manufacture its moat out of formulation, delivery, and paperwork. Whether that holds up against competition for fifteen years is a question the science doesn’t touch and the business cannot yet answer.
The most telling skeptics are the companies with the deepest pockets and the most to gain. With one big exception, the giants of the drug industry have watched this field from the sidelines instead of buying in. The exception is Johnson & Johnson, and it arrived through a side door. Its product is esketamine, a ketamine cousin it turned into the Spravato nasal spray, which is not a classic psychedelic and carries none of the hallucination. The one major company to reach for the real thing is AbbVie, and where it reached is instructive. It put more than three billion dollars of potential value into two deals with the New York biotech Gilgamesh Pharmaceuticals. And that money went toward molecules engineered to shorten or erase the trip, including a fast-acting compound called bretisilocin. It did not go toward the supervised six-hour session. The caution of everyone else is itself a piece of data. It reflects a judgment that the regulatory path is unstable, that reimbursement codes for a day-long supervised drug do not yet exist, and that a treatment in which the therapist is part of the dose carries a liability the industry has learned to fear. One event would change this calculus more than any other: a clean win in a final-stage trial for a molecule that heals without the trip. If the hallucination is erased, the broken blind that has dogged every psychedelic trial repairs itself. The supervised session disappears. What remains is an ordinary pill that an ordinary salesforce already knows how to sell. That is the event the whole industry is waiting for.
Underneath the science sits a financial reality that moves on its own logic. Every company carrying these molecules toward approval is pre-revenue. It earns nothing, spends heavily year after year, and raises the cash to keep going by selling new shares. That means each existing owner’s slice shrinks with every round. Cybin, the company now called Helus, raised a hundred and seventy-five million dollars in a single recent stretch, and the pattern repeats across the sector. These stocks don’t trade on earnings, because there are none. They trade on the next trial result and the mood of the moment. A single readout, a regulatory letter, an executive order, a rumored deal, and the stock swings hard. The field has already lived through one full cycle of euphoria and collapse. A wave of psychedelic companies went public around 2021, and most of them lost the bulk of their value within two years. Anyone watching from outside has to hold two facts at the same time: the biology may be real, and the financial vehicles carrying it are volatile, dilutive, and priced on mood as much as on results.
Hold all of that against what waits on the other side, and the shape of the wager becomes clear. It is asymmetric. If even one non-hallucinogenic plastogen clears a real Phase 3, the prize is enormous. It would be a pill that rebuilds the wiring of a depressed brain, with no clinic, no monitor, and no lost day, sold into a market of hundreds of millions. And it would carry a new way of seeing mental illness, not as a permanent chemical imbalance to be managed, but as damaged wiring that can be repaired. If they all fail, the lesson is nearly as clear. The trip was the point after all. The conscious experience was doing the work, and the field folds back into a smaller, supervised, therapy-bound niche. A disciplined observer doesn’t pretend to know which way it breaks. The move is to treat it as optionality. A defined and survivable downside on any single name, set against an upside that, if the platform is real, is hard to even put a number on.
So watching this field means watching the right handful of events and ignoring the noise in between. Three sets of catalysts arrive on a near horizon. The first is commercial. Once Compass’s psilocybin clears the agency, the real question is whether a six-hour supervised session can actually be reimbursed in the real world, the way the Spravato clinics eventually managed after their own slow start. The approval is not the test, but the first insurance payments are. The second is replication at scale, due through 2026 as the trip-intact molecules report their late-stage trials. The LSD tablet known as MM120 reads out across the year in both anxiety and depression, and the longer-lasting deuterated psilocin pill is close behind. These answer whether the classic effect survives large, rigorous study. The third is the one that matters most, and the one to watch hardest. It is the first controlled efficacy signal from a molecule built to heal without a trip, and Delix’s at-home pill is furthest into that test. A convincing result there would do more than lift one company. It would convert the entire plasticity thesis from a hypothesis into a fact, and tell a careful observer that the bet has begun to pay.
Step back from this list of risks and notice what the inversion did not do. It did not break the thesis. It set its limits. Every danger named is real, and not one of them touches the finding that started the whole race. A few doses can do in weeks what daily pills taken for years cannot, and the change is structural, written into the wiring instead of topped up in the bloodstream. That finding is not in dispute. What remains in dispute is how far it reaches and who gets to sell it. Those are the best kind of open questions, the kind with trials attached and dates already on the calendar. Within roughly a year, a compound dismissed as worthless and dangerous for half a century is likely to become prescription medicine. A category that did not exist is about to be handed its first billing codes. And the oldest question in psychiatry, whether a broken mind can be physically rebuilt rather than merely managed, has stopped being a matter of philosophy and become a measurable endpoint in a controlled trial. The molecules may yet disappoint, and the trip may turn out to be essential after all. But the brain’s willingness to be rewired in adulthood, the single discovery beneath all of it, is no longer a mystery waiting to be argued over. It is a target. And it is being shot at right now by more capital, more science, and more regulatory goodwill than the field has ever had aimed in one direction. Whatever the next readouts say, something real is being built, and it is being built in the open, on a clock anyone can watch.