Matthew Johnson on loosening the priors that run a human life

After the ads, Fridman asks for a whirlwind: what are psychedelics, and how should they be grouped? Johnson starts broader than chemistry. In common use the word covers compounds that work in different pharmacological ways. The so-called classic psychedelics — psilocybin and related compounds in mushrooms, LSD, DMT (smoked or in ayahuasca), and mescaline in peyote and San Pedro cactus — all act as agonists at a subtype of serotonin receptor, the serotonin 2A receptor. PCP, ketamine, MDMA, and ibogaine get called psychedelics too, he says, but they work by very different routes and have different effects, even if the subjective overlap is large enough that people lump them together.

The overlap that matters, compared with caffeine, cocaine, or a sleeping pill, is that they strongly alter a person’s sense of reality, including the sense of self. Cannabis, he adds, was often called a minor psychedelic in the 1970s, and under that second definition it still fits — especially if the person has no tolerance. Chemistry titles in papers are less useful here than receptor activity. All of this, as far as science can currently say, starts at the synapse.

He uses a baseball picture. One neuron is the pitcher, releasing transmitter into the gap; the other is the catcher, holding a mitt of receptors. Classic psychedelics slide into the same catcher’s mitt serotonin uses, then push a different signaling pathway inside the cell. MDMA works on the other side. It floods the gap with the brain’s own serotonin, as if the pitcher suddenly started throwing a ball every second. That is why grouping these drugs by the feeling they produce is easier than grouping them by the molecule on the bottle, and why MDMA should not be treated as a slightly friendlier version of LSD.

At the right dose, Johnson says, the classics have very strong effects on one’s sense of reality and of self. What sets the major examples apart from MDMA and the rest is physiological safety. For LSD and psilocybin there is no known lethal overdose unless someone already has severe heart disease, because the drugs modestly raise blood pressure — the same person might have a cardiac event shoveling snow or climbing stairs. For most people, a thousand times an effective dose will not damage organs or stop the brain from breathing. He would never call any compound simply “safe,” but these are “freakishly safe at the physiological level.” Aspirin is not like that. Caffeine is not like that. Most drugs kill you, or seriously injure you, at five, ten, twenty, maybe a hundred times the effective dose.

Fridman hears a duality: a hell of a journey in the mind, and almost no chance of a lethal bodily effect. Johnson agrees, then draws a contrast with exercise and with other drugs. To get those intense levels from running you need persistence; with a proper psychedelic dose, “they’re strapped into the roller coaster, and something interesting is going to happen.” In his lab he works with caffeine, alcohol, methamphetamine, cocaine. You cannot safely give a regular cocaine user a dose they would call the strongest they have ever had. With screened volunteers, you can give psilocybin or LSD at a physiological level that has a very good chance of being the most intense psychological experience of that person’s life, and for most people zero chance of killing them.

The real risk he names is behavioral toxicity: a fancy way of saying doing something stupid while extremely intoxicated — wandering into traffic, falling from a height, the same class of harm people find on high-dose alcohol. The other striking fact is that classic psychedelics are not addictive, which is almost unheard of among so-called drugs of abuse. Caffeine, alcohol, cocaine, cannabis: people slide into daily use and then into a self-control problem. With the classics, the more typical story is building up the courage to do it, hitting tolerance, or taking acid seven days in a row on a college dare. That is not the voice that says, I have to stop drinking every night.

LSD, Johnson notes, is close to nature but not quite: ergot and morning-glory seeds carry near relatives, so it is semi-synthetic. Once the conversation leaves the classics, the prominent compound is MDMA, also called ecstasy or Molly. Place cocaine at one end of an addiction continuum and classic psychedelics at the other, and MDMA sits toward the classic side. Daily use is rare, but a once-a-week pattern that is hard to stop does happen. It is also more physiologically dangerous. Therapeutic work on PTSD looks promising on ordinary medical risk-benefit terms, yet a certain dose and frequency can cause long-term damage to the brain’s serotonin system, and the cardiovascular load is heavier. You cannot push the dose as high, even in a research setting, as you can with psilocybin or LSD.

The subjective difference matters more than the safety margin. MDMA is not the full journey. Johnson calls it a psychedelic with rose-colored glasses on, more of a heart trip than a head trip. Reality does not unravel as often; the perceptual system still works; the person is not completely detached. Ego loss and transcendental unity are rare. People still experience themselves as a self, but with warmth and empathy for humanity and for the people they love. Those effects show up under classic psychedelics too, he says, as a subset. Draw Venn diagrams and MDMA’s whole circle sits inside the classic one: a psilocybin session can look just like an MDMA session, and it can also go somewhere MDMA does not go. The reverse is not true.

Asked for one general thing classic psychedelics do to the mind, Johnson says they expand the range of experience. People come out saying they did not know an experience like that was possible. Questionnaires can turn reports into numbers, but they never point to a ground truth: if someone felt they touched another dimension or sensed the reality of God, that often says more about the priors they brought in than about a quintessential mechanism. Plenty of people keep a fully naturalistic view and still have profound, helpful sessions. For some, the experience shifts them toward idealism, toward the idea that reality might be more about consciousness than matter. Science, he says, has essentially zero to say about validating those claims. What it can study is the expansion itself.

Fridman wonders how the language you bring — God, or a surface reading of quantum mechanics and spacetime — colors the journey. Johnson thinks very much so, and he worries that some of the science is overstating belief change in particular domains. The better picture, still in its infancy, is that these drugs tap general psychological mechanisms by reducing the influence of priors. Adults ride on heuristic after heuristic; evolution shaped that efficiency, and it comes at a cost. The session seems to allow greater mental flexibility and openness, so a person can be less captured by assumptions and still have the content of the trip dyed by whatever they have been reading. A philosophy professor in a high-dose psilocybin study told him, in Johnson’s telling, that it was like Hegel’s opposites: a thing he had thought about for years, and now he actually got it.

The same pattern shows up in the cancer patients and the people trying to quit smoking. They had told themselves a million times to quit, that they were ruining their lives, that they should not let the illness defeat them. They knew it in the head. After the session they felt it in the heart. That is why Johnson is both interested in and cautious about first-principles innovation for engineers. There were early, methodologically loose studies in the late 1960s and early 1970s; architects and engineers brought problems they had been stuck on for months and took about 100 micrograms of LSD, a modest enough dose to keep working. Anecdotes pile up — Silicon Valley lore, shareware, Kary Mullis saying he would not have invented PCR without psychedelic experience. “We know really next to nothing. There’s no rigorous research on that.” His gut prefers a high-dose journey you cannot sit in front of a computer for, with the breakthrough in the integration afterward. Animal work shows forms of neuroplasticity in the days after; whether that is related to therapeutic effects in humans is unknown. The study he actually wants is a hundred scientists stuck on a problem, randomized against a credible active control, plus the same canned experimental psychology tasks for everyone so you can talk about mechanism instead of face validity.

Therapeutically, Johnson is most interested in psilocybin and LSD — the modern era has mostly used mushrooms, and he had just received a grant from the Heffter Research Institute to start an LSD study. If the question is the nature of a completely altered sense of reality and self, he points to high-dose smoked, vaporized, or intravenous DMT, and to 5-MeO-DMT. People who know the terrain, including Terence McKenna and Joe Rogan, will say those are an order of magnitude beyond even a high dose of psilocybin or LSD. Therapeutic potential is less of a sure bet, because the person is not contemplating their life so much as watching reality rip at the seams. The striking pharmacological fact is that, unlike anesthetics or high-dose alcohol, the classics can throw you that far out while you remain aware and able to remember.

A common smoked-DMT method takes about three big inhalations for what people call a breakthrough: not merely a radical psilocybin-like state in this reality, but somewhere else. The first exhalation, a few seconds in, already feels like a 30-milligram psilocybin or 300-microgram LSD trip. The second hit may already be the strongest psychedelic experience of that person’s life. Johnson’s physics metaphor is acceleration, not speed: receptors filling, zero to sixty in Tesla time. If they can take the third hit, some people are propelled into another reality. Survey work finds “entities” of many types; elves tend to pop up. He strongly presumes the coloring is cultural — mechanical selves dribbling basketballs, little animals, aliens — but the underlying report is of being surrounded by autonomous, intelligent beings. People come back with communication that feels self-validating, like a download from a higher dimension, and almost never with a concrete claim you could check.

From a naturalistic angle, Johnson’s speculation is Jungian without requiring the entities to be out there. Comprehension may be so far beyond the usual that the nervous system can only orient by metaphor. After millions of years of evolution, certain characters and stories may be wired in; the narrative self is itself a story, which is why a dying mother in a psilocybin session still beats the best data when you present to Congress. At another level, beyond the receptor, the brain starts communicating with itself in a massively different way, with areas that do not normally talk. Casting the net that wide brings insight and novel thought, and it also brings false positives. When Fridman raises Rogan’s story of entities mocking him into not taking himself too seriously, Johnson remembers first noticing Rogan about 15 years earlier as another person obsessed with both DMT and Brazilian jiu-jitsu. Rogan’s tellings resemble McKenna’s; Johnson’s guess is that hearing the poet colored the later trip. Poets give you words. In the research, they do not call these episodes bad trips. They call them challenging experiences.

Johnson’s main lens on addiction is behavioral economics: take microeconomic tools built for markets and apply them to non-economic behavior. Psychologists noticed that another discipline had been studying behavior all along, just while staring at spending and saving. He has run demand-curve analyses on cigarettes and nicotine products — how much someone will consume as “price” changes, including how many times they will pull a knob in a three-hour lab session after six hours without a smoke. Elasticity is the key idea. Gasoline is often inelastic: the price doubles and consumption may fall only 10 percent. Push the price to a hundred dollars a gallon and the curve eventually turns. Reward is not one-dimensional. At a near-zero price, diamonds look like the better good; keep jacking the price and you keep buying toilet paper. Which reward is “better” depends on the price.

That is one half of his definition of addiction, which he does not think maps cleanly onto the DSM. A couple of beers is not a problem. It becomes one when that reward outcompetes work, marriage, and the rest of a prosocial life. The other half is delay discounting, or time preference: postponing a good damages its value. Everyone would rather have things now than later. Ask “nine dollars today or ten tomorrow” and you see large differences between addicted and non-addicted groups — not only heroin and cocaine users, but everyday cigarette smokers — even when the reward is money, not their drug. Map the curve at a day, a week, a month, a year, ten years, and the shape across pigeons, rats, and humans is hyperbolic, not exponential. The first unit of delay hammers value; later units hurt less. Exponential discounting, the logic of compound interest, cuts the same proportion every period. Hyperbolic curves produce preference reversals.

A year from now, nine dollars in 364 days versus ten in 365 looks like free money; bring the same pair to today versus tomorrow and many people grab the nine. Nothing about the options changed except proximity. Johnson’s image is two agents sharing a bus: one driver wants the long term, the other wants now. In the doctor’s office you are done with the drug; that evening a friend has scored and the reward is in front of you. The snooze alarm is the everyday version. So is the snack that is thirty seconds away in the fridge rather than at the grocery store. Heavy cigarette taxes do cut youth smoking, but for the poorest inelastic smokers you are taxing their existence and stripping the money that might have competed with the habit. Dealers, legal or not, sense the same curve: lose some customers, more than make it up on the rest. A drug-war goal of making illegal prices as high as possible, without that nuance, can make a poor family poorer and, after a felony, make drugs a better remaining reward.

Asked whether a world of fully legal drugs would be safer, Johnson says it depends what legalization means. Almost nothing is laissez-faire: even caffeine is capped in sodas and over-the-counter pills. Alcohol is illegal at 20 and illegal in a moving car. Psilocybin is Schedule I, yet methamphetamine and cocaine, ironically less restricted as Schedule II, are legal in licensed human research. He would rather talk about regulating all drugs than about a binary. He would not put heroin and cocaine in convenience stores. Safe injection sites make sense to him: if people are using anyway, do it where the water is not from a puddle, where overdose can be reversed, and where a professional — unlike a clerk — can teach harm reduction. With naloxone on hand, a medical professional can revive almost every opioid overdose; if every current death happened in that room, he says, there would be almost none. Switzerland has gone further and provided the drug itself. The “wrong message” argument, in his view, does not survive contact with daily shattered families, or with the fact that a felony for a 19-year-old’s psilocybin mistake is a scarlet letter on employment and student loans. He had just testified in New Jersey to move that penalty toward a misdemeanor; two committee votes still worried about the message.

Then Fridman asks which drug is the most addictive, the most dangerous, in society today. Johnson’s answer is the uncomfortable one. If the metric is mortality, it is nicotine. Tobacco kills about half a million people a year in the United States and five to six million worldwide, four times alcohol in the U.S. Graph alcohol, tobacco, and every illegal drug together — opioids, cocaine, meth, the rest — and the illegal pile is a barely visible blip. He calls it the kind of truth you do not know what to do with, in the same family as terrorism versus swimming pools, or a society frantic about coronavirus that is not frantic about cigarettes. He is not advocating making tobacco illegal. That, he thinks, would be a horrible mistake.

The path he can imagine is a mandated drop in cigarette nicotine, but only in conjunction with fully regulated e-cigarettes as a much safer, not safe, alternative. The unique fact to exploit is that nicotine is largely not what causes the harm; the aromatic hydrocarbons and carcinogens in burning tobacco smoke are. This is not alcohol prohibition, where near-beer cannot meet the demand. Tax cigarettes to the sky, ban flavors, then over-regulate and overtax the alternative, and you push people toward a black market in real cigarettes — the same loose-cigarette trade that, he notes, led police to Eric Garner. He has some hope for a thoughtful version of this and little confidence it will be executed that way. Nicotine and tobacco, he fears, are in danger of replaying every other failed prohibition, just as the larger ship is slowly turning away from that model.

Johnson has also put delay discounting onto sexual decisions, across roughly twenty papers and several drugs. Volunteers who already use cocaine — never never-users, never people trying to quit — take the drug in a medically supported lab. They look at photographs, pick people they most and least want a one-night stand with and whom they judge most and least likely to carry an STI, then read a vignette: this person had casual sex with you, a condom is available, rate from 1 to 100 whether you would use it. At zero delay, most people say they definitely would, often 100. Then the wait becomes five minutes, an hour, three hours, twenty-four hours: sex now without a condom versus later with one. Under the hood, sex without a condom often feels better; condom use is still the larger later reward because it keeps most of the benefit without the concurrent risk of pregnancy or infection.

Alcohol research had shown little effect on monetary delay questions, which struck Johnson as unbelievable given how drunk people actually behave. Risky sex, not reallocating a retirement account, is the typical context. In his administration studies, cocaine, alcohol, and methamphetamine do not shift the zero-delay item — if the condom is right there, people largely say they will use it. The effect appears once delay is in the model. A survey of men who have sex with men found that about a quarter reported, in the previous six months, unprotected anal intercourse in a situation where they would have used a condom and simply did not have one. Most of the action, he argues, unfolds under those suboptimal conditions: things get heavy, nobody wants a ten-minute walk to a closed store. Help is mostly on the environment side. Carry one if you are going to drink or use stimulants. Make condoms widely available. He would like to try contingency management — the same operant method that is among the most powerful ways to cut cocaine use, paying for a clean urine — by texting a young single person every few days: send a photo of a condom within a minute, get five dollars. You are not removing the main reward. You are making the safer version available.

He also sketches two businesses he does not plan to open. One is five-minute condom delivery anywhere, a DoorDash for that delay. The other is a two-driver ride home: you meant to have one or two drinks, you had five or six, an ordinary Uber leaves the car to get towed, and some people take the risk. Two sober drivers arrive; one takes you home in your car, the other follows. It might cost fifty or sixty dollars, more than a regular Uber, but the night at the bar already cost that, and the car is in the driveway in the morning. Stimulants and alcohol, he thinks, hit different levers. Cocaine and meth raise the rewarding value of sex itself — users say sex on those drugs is so much better that quitting threatens the sex life — and his group ran what he describes as the first double-blind lab demonstration that cocaine raises sexual desire even from pictures. Alcohol works more through disinhibition: it weakens the grip of a punisher that will arrive tomorrow, next week, or in twenty years. A tabloid headline reduced the condom paper to “horny cocaine users don’t use condoms.” That, he admits, is roughly what it found.

Johnson’s favorite psychedelic finding is the smoking work, chosen because tobacco is such a serious addiction and because a behaviorist can verify the change. Mystical reports interest him; urine and exhaled carbon monoxide are where the rubber meets the road. The pilot enrolled 15 people, not psychonauts. Most were not particularly interested in psychedelics. They had tried everything, they wanted to quit, and Hopkins sounded safe enough. That, he likes, is a better model of a future average patient. They used pure psilocybin in a capsule, moderately high and high doses — about 30 milligrams, body-weight adjusted up toward 40 or even 50 for heavier people, roughly five dried grams of *Psilocybe cubensis*, the “heroic dose” McKenna talked about. This is not a museum or concert amount. Even experienced volunteers have told him they had not known the orbit went that far.

Preparation is long: on the order of eight hours with the two guides who will sit in the room, building rapport, teaching how to handle fear, and running ordinary cognitive behavioral therapy for smoking — a target quit date that is also the first drug day, a smoking diary, pluses and minuses, how to treat a slip as data instead of a reason to collapse. Session day is a living room more than a lab. They swallow the capsule, lie on a couch for five or six hours with eyeshades and headphones, mostly classical music, though a later paper found Tibetan bowls and gongs no worse and, non-significantly, a bit better. The so-called guides are a safety net. Blood pressure is checked; a physician is on hand; the emergency department is across the street; the DEA has approved the work. The instructions are trust, let go, be open: if there is a door, open it; if there is a monster, walk toward it. McKenna’s comparison, which Johnson quotes, is that meditation is years on the accelerator, and a psychedelic is a fully loaded semi speeding down a mountain, and your job is not to slam the brake. People come out exhausted. Some call it the hardest thing they have done.

The pilot was small and not randomized; the only honest question was whether follow-up was worth it. Biologically confirmed abstinence was 80 percent at six months and 60 percent at an average of two and a half years. The best published smoking-cessation packages sit in the upper 50s, and those use two medications for months plus weekly or near-weekly therapy for a year. The Hopkins pilot used three psilocybin doses inside a 15-week course. The randomized trial now underway scaled that to one session versus the FDA-standard nicotine patch, 40 people each, same CBT, with fMRI before and after. Among the 44 who had reached one-year follow-up, 59 percent of the psilocybin group were biologically smoke-free versus 27 percent on the patch — and the patch arm was itself unusually good. Numbers could still move. Volunteers need to be in commuting distance of Baltimore. Fridman, imagining hundreds of thousands of listeners, asks how to reach him. Johnson’s answer is not his inbox. Go to the Johns Hopkins psychedelic center site, or search clinicaltrials.gov for psilocybin; do the online screen. He is already failing to answer colleagues, mentees, and parents. He is hiring a postdoc. He has a three-year-old on the floor with blocks, and some emails will never get a reply.

What has this work done to Johnson’s picture of the mind? As a behaviorist he treats “mind” as a metaphor for behavior, but the first change is an appreciation for the magnitude of human experience. His working assumption is that every phenomenal moment has a biological side — serotonin, dopamine, network communication among brain areas. Prompt someone to think of a loved one and something biological happens. Experience off drugs is already a controlled hallucination, a simulation. You cannot prove a negative, but the decent scientific guess is a one-to-one correspondence: every psychological event has a biological correlate. That is why it still floors him that you can throw a relatively small number of molecules into the nervous system of a 60-year-old who has hiked Everest, speaks five languages, has grandchildren, and hear that person say this fundamentally changed who I am and what I think life is about. Paperwork and dud sessions dull the feeling. Then someone gives the Full Monty report, and he remembers why they do the work.

It also fills him with the sense that we have lit a match in a dark room. Humans always think they have 99 percent figured out and are polishing the last 1 percent; he wants the reminder that we know almost nothing. The future of “drugs,” if the species lasts, might be molecules aimed at a subset of serotonin 2A receptors in a particular cortical layer, in sequence, combined with magnetic stimulation and a baseball cap doing what now takes giant equipment, combined with virtual reality that moves from out here to in there. External VR and the internal VR of perception could meet. He is willing to sound foolish: maybe most of life migrates into internal worlds; maybe a regular lifespan contains a thousand years of experience in two seconds. On Neuralink and brain-computer interfaces he is broadly for it, for the same reason he is for psychedelic research, with the same caveat that a brave new world has a dark ethical side. In the short term, helping serious neurological disorders is an easy yes. He is sympathetic to Elon Musk getting arrows for oversimplifying in public, because scientists are bad at making work meaningful without being belittled, and because first-principles claims that annoy every lab this year can still be the 20-year path. Aggressive timelines are a form of delay discounting: aim at a year, miss, and you may still finish in three; aim at twenty and you may never start.

Johnson had recently written a paper whose title is the warning: “Consciousness, Religion, and Gurus: Pitfalls of Psychedelic Medicine.” He separates what he can say as a scientist from armchair speculation. Phenomenal consciousness — Nagel’s what-it-is-like, the hard problem — still strikes him as a real problem, not a confusion to dissolve. By and large, skeptically, psychedelics have not taught us anything about that. They have been used around other things called consciousness: the moment-to-moment experiential self, and the narrative self that strings stories together. A popular story said psilocybin decouples the default mode network, a set of regions including posterior cingulate and medial prefrontal cortex tied to self-referential processing, and that this is ego dissolution. Johnson thinks that story left the station too fast. More and more drugs decouple that network, including ones that are not psychedelic. It may just be a signature of being out of your head. They found DMN decoupling on salvinorin A, a smoked, non-classic compound that often involves entities and does not produce the clear ego-loss people report on classics. Other networks change too. Self-referential processing in the DMN was already known before this wave of work.

What remains striking is that people report unity experiences with high reliability: boundaries of the self lost, minimized, or expanded until there is no edge. That, he thinks, is useful for understanding the sense of self. It is not the hard problem. You can build an AI that passes a Turing test for selfhood, that talks and behaves as if an agent is running the show — an illusion that may even be computationally useful — and that demonstration is silent on whether there is an experience inside. Fridman has been struggling with the words. He used to say there is no difference between the appearance and the existence of consciousness; what he means is that as the portrayal gets more realistic, the gap between real and fake is smaller than people like to imagine. Fake it until you make it, far enough, and you look back and the system is the thing. Philosophers hate that, he says; it is closer to an engineer’s language. Johnson is a radical empiricist, not a materialist by dogma. What we do today would have been magic five hundred years ago; the boundary of nature moves. He is open to so-called psi phenomena as possibly another layer of nature, and he is attracted, as speculation he cannot say as a scientist, to panpsychist and Russellian-monist ideas: that the inside of being a particle, or spacetime itself, is a fundamental, and that human consciousness is a heavily processed derivative. Occam’s razor, for him, strains at the moment complexity magically grows an inside out of a purely objective universe of interactions.

Then Fridman asks the craziest question: could DMT entities be extraterrestrial life communicating through chemistry rather than ships? As a good scientist, Johnson says, you first have to say it is possible. The interesting part is the probability — one in a million, one in a trillion, one in more than the atoms in the universe — and a test. Bring back a verifiable fact about another planet, or a prediction we will be able to check in five, ten, twenty years, the way some of Einstein’s work waited decades. It is also possible such beings would not share our appetite for prediction, would not sign a consent form, would not care about our experiments. Rick Strassman, who ran high-dose DMT work in the 1990s, labeled as speculation the idea of entities from another dimension. Johnson thinks science is too often closed, even punitive, about 20-year thoughts, because the next grant and the next paper do not care. The practice he likes is the one he also hears in Fridman: most of the week, rigor; every so often, step outside and talk about aliens, consciousness, free will, wormholes. Newton spent a great deal of time decoding the Bible. The greats did not treat “I am a scientist” and “I am trying to figure out the world” as two jobs, and they did not observe the silos.

Fridman asks whether Johnson is afraid of death, and whether the research has changed that. He ponders it probably daily. He is afraid of dying, the process; he is not afraid of being dead. He cites Penn Jillette, who may have gotten it from someone else: he is not afraid of 1862 before he existed, and he is not afraid of 2062 after he is gone. The dying itself is usually not good. He would rather fade with old age, conscious, many years from now, than go in an accident or a horrible disease he has already watched in people he loves. He is more concerned about not being there for a three-year-old son, a wife, a mother. Zoomed further out, he thinks maybe too much about whether the species makes it off the planet. Nuclear weapons, pandemics, a future in which almost anyone can build something a million times worse than COVID, climate: the only real guarantee of another million years is a diversified portfolio. Stay here if we can, but get a foot off the world. A colony on Mars by the time he is 120 would comfort him. It would not solve death. He would still not be there for the rest of the observable universe, and that, he agrees with Ernest Becker’s line of thought, is the terror we mostly live by denying — the illusion that this goes on forever, when it might end today.

Two things have moved him toward acceptance: being with people close to him as they died, and the cancer study, in which psilocybin substantially helped the vast majority with existential distress. At some point the stars go out; the era in which life is possible is a blink. If, at the end, he could choose a universe with some unknown afterlife versus one with nothing, he would want to see what is behind the door. He is not going to kill himself to look; if the button were there, a true scientist’s curiosity would reach for it. He also knows the real test is the deathbed, or a terminal diagnosis. Easy to say now. Then Fridman asks whether any of this has meaning. Johnson has been doing the work almost 17 years. He remembers a healthy volunteer, in no treatment study, contemplating the suicide of his son — as intense and vulnerable as human experience gets — and feeling lucky to be trusted that far. Meaning, for him, is defined meaning: the beauty despite the absolute ugliness. We come from filth. We are animals. We are descended from murderers and rapists. Against that background we still manage self-sacrifice, connection, truth-seeking, music, art. That possibility is the meaning. The dark parts belong in the same account. So-called bad trips are challenging experiences; the death of a loved one is among the greatest learning experiences, which we try to minimize and can still treat as richness if they happen. Psychedelics, art, science, and a society that expands what can be experienced are forms of the same move. He can accept that it all ends, and even find it beautiful *because* it ends, because it came from nothing, and because beauty and truth and connection still happened. He hopes he still believes that when he is the one in the bed.