What this covers

Theodore Schwartz, a neurosurgeon, walks through what a career spent operating on brains reveals about free will, consciousness, and the messy reality of medical practice. The conversation ranges from how tumors and electrical stimulation can create or erase beliefs and desires—suggesting personality is a product of circuit function rather than an immutable self—to how the storyteller brain constructs a unified narrative after the fact, making us feel like authors of decisions we did not consciously make. Schwartz draws on decades of surgical cases, neuroscience evidence from split-brain patients and seizure disorders, and philosophical thought experiments to argue that free will is largely illusory, even as he acknowledges the paradoxes this creates for how we assign responsibility and understand identity.

Beyond philosophy, Schwartz examines the ungoverned craft of neurosurgery itself. He describes how outcome measurement is sparse—surgeons' individual skill levels are unknowable to patients, reimbursement ignores experience, and the profession accumulates almost no systematic data on who performs best—leaving patients to rely on word-of-mouth and multiple opinions rather than transparent quality metrics. He explores emerging tools like brain-computer interfaces and focused ultrasound that may eventually permit surgery without opening the skull, and recounts the history of Harvey Cushing reducing operative mortality, the training paradox that residents must learn on real patients or never be ready, and the tension between confidence and humility required to earn trust while staying honest about mistakes. A tangent on the JFK autopsy and a sustained meditation on how seizure-induced religious experiences in Dostoyevsky vanish once the epilepsy is treated illustrate his broader claim that belief itself is substrate-dependent.

Sharpest takeaway

Schwartz argues that the self and free will are stories the brain constructs after the fact, since any aspect of personality, belief, or emotion can be altered by changing brain circuitry — and that neurosurgical practice itself reveals how messy, unmeasured, and human the craft of operating on the brain remains.

  • Split-brain and hemispherectomy patients still feel like a unified self despite divided or halved brains, undermining the unified-self model
  • Deep brain stimulation, tumors, and seizures can create or extinguish desires, beliefs (even religious experiences), and personality traits
  • Neurosurgical quality is largely unmeasured and reimbursement ignores experience, so surgical decisions hinge on individual skill that systems do not track

The argument · threads10 threads · 41 claims
0.85

The brain's circuitry, not anatomy or unified self, determines all behavior, personality, and identity.

8 pointscentrality 5/5
  • All that we are is our brain, which contains hundreds of different modules processing information and making decisions that we then carry out, plus another part of the brain that acts as a storyteller, constructing a logical timeline after the fact that makes us feel a unified self made the decision — when in fact that self did not really exist.

    “there are hundreds of different modules processing information in the brain and determining what behaviors we should do and making decisions for us, and then we carry out that behavior. Then, there’s another part of the brain that basically is a storyteller, that tries to make sense of it all after the fact.”

  • Behavior is so dependent on brain circuitry that a documented patient developed an uncontrollable interest in child pornography as a frontal-lobe tumor grew, lost those desires when the tumor was removed, and had the desires return years later when the tumor recurred — demonstrating, like frontal lobotomies, that the frontal lobe is highly sensitive for behavior and that personality is a function of circuit health.

    “as the tumor got bigger, pushing on their frontal lobe, they became very interested in child pornography and couldn’t control themselves... they took the tumor out, and those desires went away. Then five years later, they came back, and they re-imaged the patient’s head, and the brain tumor had recurred.”

  • Epilepsy is a disease of the brain's circuitry rather than its anatomy (the brain often looks completely normal during epilepsy surgery), and some seizure symptoms mimic profound human experience — Dostoevsky experienced the certain presence of God during seizures, and such religious experiences disappear once the epilepsy is cured by medicine or surgery — showing that beliefs and religious experiences can be altered by altering the brain, which tells us much about religion and the human mind.

    “Dostoevsky, who was an epileptic. He used to experience the presence of God when he had a seizure... when you treat them... and you cure them of their epilepsy, their religious experiences go away.”

  • The unified self does not exist in the way people think, as demonstrated by neurosurgery: a corpus callosotomy splits the brain into two independently functioning halves controlling each side of the body in opposite ways, yet the patient still feels like one unified self, and a hemispherectomy removes half the brain without the patient waking up feeling like half a person — showing the brain creates the sense of a single self despite a divided or reduced substrate.

    “there’s a neurosurgical operation called a corpus callosotomy, where we can split the two sides of the brain in half so that they become two independently functioning brains, and that person does not feel like two people.”

  • Because every behavior, desire, and feeling derives from brain circuits firing in particular ways, deep brain stimulation can create or extinguish them — exemplified by a patient who liked a particular musician only while the stimulator was on — and therefore any part of your personality can in principle be altered by placing electrodes in the right locations and changing your brain's circuitry, just as Alzheimer's changes wiring and makes someone a different person.

    “any part of your personality can be altered with an electrode or a number of electrodes placed in the right places and altering the circuitry of your brain appropriately.”

  • We do not have free will in the way most people believe: our brains make decisions for us, we carry out the resulting behaviors, and then we construct a story that arranges it into a logical timeline making us feel a self made the choice; the storyteller part itself also lacks free will, and we do not know where our ideas come from — they simply emerge into consciousness, which itself remains unexplained.

    “I do not think we have free will in the way that most people do. I think that our brains make decisions for us. We carry out those behaviors, and then we write a story that makes it into a logical timeline that makes us feel as if we were the ones”

  • Freud is probably underrated circa 2025 because his core insight — that an unconscious controls what we do while the ego is a tiny thing, a revolution comparable to Copernicus and Darwin in stripping us of self-control — is vindicated by modern neuroscience showing the brain does enormous processing we are not conscious of; while his specifics like penis envy and the Oedipal complex invite criticism, his structure of the mind subverting the ego and validating the unconscious was brilliant.

    “Now, circa 2025, we realize that the brain is doing an enormous amount of processing that we’re not conscious of. That unconscious processing is exactly what Freud was getting at back in Victorian Vienna.”

  • Personal identity is in constant flux — the Tyler Cowen of five seconds ago differs from the one sitting here now — so if you replicated a person into two copies, those two would immediately diverge into completely different people one second later because their molecules, experiences, and locations in space would differ; on this reasoning Schwartz would step into a Star Trek transporter.

    “If you could replicate yourself so that there were two Tyler Cowens, I think those two people would diverge in who they are. They would be completely different people one second after they became a different person because their molecules would be on a different path.”

0.74

Neurosurgical excellence requires maturity, wisdom, and character traits that cannot be taught quickly.

5 pointscentrality 5/5
  • Becoming a great neurosurgeon requires not just accumulated operative experience but maturity and wisdom that come with age, which is why fully compressing the training timeline is risky even if the technical skills can be taught faster.

    “It’s not just years that you’ve been doing it, but it’s also wisdom. Wisdom comes with age as it does with experience.”

  • Neurosurgeons reportedly carry more ego and confidence than even Wall Street traders, but excellence requires holding confidence and humility together: enough confidence to earn a patient's trust to do risky work, and enough humility to introspect honestly after the inevitable mistakes and ask how to do better — because excessive confidence becomes arrogance that lies to patients about perfect outcomes and forecloses the self-examination needed to improve.

    “you have to, at the same time, have confidence and humility. You have to combine the two, and it’s very difficult to do.”

  • Raw IQ matters less for neurosurgical excellence than commonly assumed — a British Medical Journal study comparing the IQs of neurosurgeons, rocket scientists, and average professionals found them all roughly the same — and what actually distinguishes great surgeons is grit, physical stamina, judgment, coordination, dedication, and the ability to maintain a positive mood under sustained pressure.

    “I quote this article that came out in the British Medical Journal that looked at the IQs of neurosurgeons, rocket scientists, and average professionals and essentially found that they were all about the same.”

  • Unlike internal medicine, neurosurgery rarely involves diagnostic or radiographic dilemmas because patients arrive already knowing their diagnosis and location from an MRI; the real difficulty is physical decision-making in the operating room, which is why AI's strength at interpreting scans does not translate into replacing the surgeon.

    “neurosurgery is different than internal medicine. We’re usually not presented with diagnostic dilemmas or radiographic dilemmas. The dilemmas that we face are physical decision-making in the operating room.”

  • Undergraduate liberal arts and medical school could be condensed from eight years into about six because much of the early curriculum (e.g., reading Nietzsche, two years of rotating to choose a specialty) is unnecessary for someone who already knows they want to be a neurosurgeon, and the actual surgical skill could be learned in about four years.

    “I do think that undergraduate education and medical school could probably be condensed into six years instead of eight.”

0.68

Emerging brain technologies promise safer, less invasive access and interface capabilities over the next fifteen to twenty years.

7 pointscentrality 4/5
  • Neuralink's distinctive advantage is that its electrodes, which record from individual neurons inside the brain, are malleable rather than rigid, so they move with the constantly pulsating brain and avoid forming the dense scar that previous rigid electrodes provoked (which tends to push electrodes out); this is one of two device classes, the other placing electrodes on the brain's surface, and it's unknown which will prove better.

    “the electrodes that he put in are malleable. They’re not rigid electrodes. The brain is constantly pulsating and moving, and the electrodes can move with the brain”

  • Electrodes placed on the surface of the brain can yield the same information as electrodes inserted into the brain, and crucially do not damage the brain, making surface BCIs the appropriate approach for healthy people (versus Neuralink's intracortical electrodes, which are placed in the movement-control region of already-paralyzed patients); electrodes can also be removed, so getting them on and off the brain without damage is already feasible, the main risk being opening the skin and skull.

    “You can get the same information from an electrode on the surface of the brain as you can from electrodes in the brain.”

  • Focused ultrasound can beam through the skull into specific brain regions to open the blood-brain barrier (which normally blocks most drugs to protect the brain from toxins), so one could foreseeably take an effective depression pill and use focused ultrasound to open the barrier only in the depression-relevant circuits, delivering the drug in high doses solely where needed — a much more effective and selective treatment than current pills.

    “you do focused ultrasound, open up the blood-brain barrier only in those circuits, in those anatomic areas that are important for depression, and that drug will only get into those specific areas in very high doses.”

  • Making the surgery to implant a brain-computer interface safe enough for routine use — to the point a healthy person could be hooked up to a future GPT — could take roughly 15 to 20 years.

    “I think that making the surgery safe could be 15, 20 years. That could happen.”

  • Brain-computer interface technology already exists in academic institutions worldwide: implants let paralyzed people move a robotic arm, drive a car, and speak by extracting language production from the brain into a computer that voices the words the person is thinking; the remaining work is making it commercially available and self-contained without external wires, and ultimately enabling non-invasive wireless thought-based communication between people.

    “If you have someone who cannot talk because they can’t move their mouth, you can take the language production out of their brain, put it into a computer, and have a robot or a picture of someone’s face speaking as that person is thinking what words they want to say. That technology already exists”

  • The biggest obstacle limiting progress in brain surgery is having to open the head with a scalpel and drill through the skull; the Holy Grail is delivering energy into the brain non-invasively — already partly achievable via focused radiation and focused ultrasound, and potentially future optical/optogenetic methods — and operating through blood vessels (e.g., treating aneurysms with a catheter via a wrist needle-stick rather than open dissection) shows how dramatically less invasive the field can become.

    “the Holy Grail for us is to put down the scalpel and to not have to drill through the skull and to be able to deliver energy into the brain without opening up the head.”

  • The brain is the key biological obstacle to radical longevity because, unlike replaceable organs such as the liver, brain cells do not reproduce or regenerate and deteriorate over time; the brain is essentially a complex information-processing machine, so the hope is that someday we could reproduce its connections inorganically and live on more permanently in another form.

    “the brain is just a very complex information-processing machine. So, the hope is that someday we’ll have some inorganic way of reproducing the same connections we have in the brain so that we can live on more permanently in another form.”

0.60

Medical reimbursement and institutional metrics fail to measure surgeon skill or patient outcomes that truly matter.

5 pointscentrality 4/5
  • Because insurance companies fix reimbursement rates, a neurosurgeon on their first day out of training is paid exactly the same amount for removing a brain tumor as a surgeon with 30 years more experience, so there is no accounting for experience or expertise in physician reimbursement even though the quality of the product can differ dramatically.

    “There’s no accounting for experience and expertise in reimbursement for physicians for what they do.”

  • Whether to observe, radiate, or operate on a tumor depends heavily on how skilled the specific surgeon being consulted is at that surgery, and because not all surgeons are equal and skill is unique to each individual, an AI cannot incorporate that decisive variable into its recommendation.

    “What’s going to factor into whether to do the surgery or not is how good the surgeon that you’re seeing is at doing that surgery. Not all surgeons are created equal.”

  • Hospital and physician priorities are only partially aligned: hospitals track readmissions, infections, and clots because that is what the government uses to judge them and set reimbursement, while the things that actually matter in a neurosurgical operation — whether the whole tumor was removed and whether the patient can see and walk afterward — are not tracked, because we measure what is easy and what regulators want rather than what is most important.

    “Unfortunately, the things we measure are the things that are easier to measure and the things that the government wants us to measure, but they’re not necessarily the most important things to be measured in a neurosurgical operation.”

  • Surgeon-level outcome data could in theory be collected by having surgeons enter results for every operation into a database, but in practice once the operating room door closes nobody is observing the surgeon, and the profession is nowhere near systematically accumulating this data.

    “Once that OR door is closed and a surgeon is on their own, there’s really nobody looking over their shoulder for the most part.”

  • Because top neurosurgeons for a given operation are known within the profession (they write the most articles, give the most lectures, do the most volume) but are hard for consumers to identify, the best strategy for a patient is to find someone in the healthcare profession to call around and ask, and to seek second, third, and fourth opinions rather than accepting a surgeon's word at face value.

    “neurosurgeons know who are the top neurosurgeons doing a particular operation because those are the people who write the most articles, give the most lectures, do the most volume.”

0.60

Neurosurgery is fundamentally a physically and mentally demanding craft requiring sustained focus and emotional detachment.

5 pointscentrality 4/5
  • The capacity to sit and practice a musical instrument for hours to reach proficiency lends itself to neurosurgery, because both require the sustained concentration and focus needed to memorize large volumes of information and to operate under the microscope for long periods.

    “the ability of sitting in a room and practicing an instrument — which I would do for hours and hours and hours — if you want to become very proficient at it, is something that lends itself to neurosurgery.”

  • Neurosurgery is a physical activity demanding the ability to stand for long periods without significant tremor, focus for hours, and function on little sleep including emergency 2am surgeries; Schwartz developed excruciating back pain early in his career and overcame it through core-strengthening exercises like planks and through practice that strengthened the specific muscles needed.

    “neurosurgery is a physical activity. You not only have to stand for long periods of time, you can’t have a significant tremor.”

  • Being cold, mechanical, and emotionally isolated from the significance of the operation can be an advantage for a neurosurgeon, allowing them to focus purely on the task and perform at a high level without being weighed down by the meaning of what they are doing; complications, though rare, occur once or twice a year for busy complex neurosurgeons and are carried with them constantly.

    “that can be an advantage, to be emotionally isolated from what you’re doing so that you can perform at a high level and not think about the significance of what you’re doing, but just think about the task that you’re doing.”

  • Neurosurgical training faces an unavoidable paradox: patients don't want residents touching them, yet the system is obligated to produce surgeons who are fully competent on the day they graduate, and because cadavers, models, and simulations are not equivalent to real operating, it is impossible to forbid trainees from doing anything while still expecting them to be ready on graduation day.

    “That’s obviously an impossible task, to not allow a trainee to do anything, and yet the day they graduate, they’re fully competent to practice on their own.”

  • Society grants enormous trust to surgeons without verifying their condition on the day of surgery — there is no drug testing or any testing of surgeons before they enter the operating room, so a patient cannot know whether their surgeon suffered a personal tragedy the night before, binged, or is impaired; while Schwartz personally abstains and sleeps well before operating, he notes this trust is given without knowing what is going on in surgeons' heads.

    “we don’t do drug testing, and we don’t do any sort of testing for surgeons before they walk into the operating room.”

0.39

Depression and epilepsy are disorders of specific brain circuits, enabling targeted interventions far more effective than broad chemical treatments.

3 pointscentrality 3/5
  • Electroconvulsive therapy works by inducing a seizure — discovered after observing that depressed patients improved following a seizure — delivered via cardiac-defibrillator-like stimulation of the skull under anesthesia; though not the most refined treatment and reserved for very refractory depression, it remains one of the best treatments currently available, working because a seizure is a powerful wave of neuronal depolarization and chemical release that alters the brain's network functioning.

    “We found out years ago that patients who had depression and then suffered a seizure got much better... right now, it is one of the best treatments we have for depression.”

  • Changing the brain's chemistry can make people more prone to risky behaviors: certain Parkinson's drugs have made patients become compulsive gamblers, and the same chemical-modification logic applies to addiction, which is why deep brain stimulation is being tried — effectively in animal models — to modify addiction, and why GLP-1 drugs appear to limit addictive behavior.

    “people who were taking a certain Parkinson’s drug, and they became compulsive gamblers. We know that certain medications that you take — changing the chemistry of your brain can make you more prone to risky behaviors in gambling, and the same is true of addiction.”

  • Biochemical correlates of depression are hard to find and current drugs cause side effects because depression likely resides in only certain frontal lobe circuits, yet a pill that alters a neurotransmitter goes everywhere in the brain rather than targeting just those circuits — so the unselectivity, not the absence of a physical basis, explains both the side effects and the difficulty of pinning down the biochemistry.

    “when you take a medication as a pill, it goes everywhere in your brain. Depression is probably only in certain places in your brain.”

0.39

Scientific progress eventually corrects institutional bias, though validation is slow and firsthand evidence justifies pioneering beliefs others reject.

3 pointscentrality 3/5
  • Schwartz's most nonconformist position came when he pioneered minimally invasive neurosurgery — operating through the nostrils with endoscopes or via small eyelid incisions instead of opening the skull — which most neurosurgeons at the time considered crazy and unsafe simply because they had never seen it done; the burden of proof fell on him to demonstrate via dozens of articles and reproducible results that the new technique was as good or better, illustrating that firsthand experience can justify a belief others reject until science validates it.

    “At the time when I was doing those operations, most neurosurgeons thought they were crazy, and that they shouldn’t be done, and that they were unsafe.”

  • Despite institutional politics and biases, the scientific method will eventually converge on the right answer and prove biased institutions wrong — but this efficiency only holds over the long run, not day-to-day or year-to-year, and the best scientist is one who states current best guesses while remaining willing to change their mind given better evidence; the food pyramid's old 'eat more carbohydrates' advice exemplifies a best guess on available evidence that later proved likely wrong.

    “in the long run, the scientific method will figure it out, and there will be one right answer. That institution — whatever their bias is — will be proven wrong in the long run.”

  • Schwartz's extraordinary output (500 articles, 200 chapters, 7 textbooks) stems from compounding advantages: operations that took eight hours now take three as skill grows, and a reputation for finishing papers attracts collaborators, so he can supply ideas and data to six or seven residents who write rough drafts he then edits — multiplying himself through people who help, an advantage unavailable at the start of a career.

    “it allows you to multiply yourself by having people who work with you on what you do.”

0.21

Neurosurgeon aging follows a bell curve with decline typically arriving in late sixties to early seventies, varying by individual.

2 pointscentrality 2/5
  • There is no absolute age at which a neurosurgeon must stop, as competence varies by individual along a bell-shaped curve, but on average the middle of that curve — where decline makes one less desirable to handle a case — falls in the late 60s to early 70s, with some surgeons remaining quite good in their early 70s.

    “I would say the middle of that bell-shaped curve is going to be late 60s, early 70s.”

  • Harvey Cushing, the founding forefather of neurosurgery around 1905, was the first surgeon to dedicate his career to neurosurgery and reduced the mortality of brain operations from about 50 percent at the beginning of his career to about 8 percent by the end — a remarkable achievement in making neurosurgery safe.

    “at the beginning of the century had a mortality of about 50 percent, and at the end of his career, it was down to about 8 percent.”

0.21

Once safe, brain-computer interfaces for enhancement should face no more medical regulation than cell phones, though societal harms warrant broader policy.

2 pointscentrality 2/5
  • Once brain-computer interfaces are completely safe, choosing to get one for enhancement should not require medical regulation any more than buying a cell phone does — a cell phone is already a slow brain-computer interface — though the uses people make of such devices may warrant regulation if they prove detrimental to society, similar to debates over regulating Facebook and X for abuse.

    “Once it’s perfectly safe, then absolutely, it will not need to be regulated any more than Facebook and X need to be regulated by the government because of abuse.”

  • On the JFK assassination, Schwartz argues for a single shooter: while the Parkland surgeon Kemp Clark did not report finding an entrance wound in the back of the head, neurosurgeon Bob Grossman, who as a junior trainee stood next to Clark, published an article 50 years later stating he felt an entrance wound there — corroborating the Warren Commission and, with one entrance wound and an exit through the parietal lobe, placing the single shooter behind Kennedy.

    “there was a second witness who was in the room at that time, who corroborated the Warren Commission’s report that there was an entrance wound in the back of the head, which then places the single shooter in the back”

0.18

Women's underrepresentation in neurosurgery stems from career demands conflicting with childbearing and family freedom preferences.

1 pointcentrality 2/5
  • Women remain underrepresented in neurosurgery (around 10 percent, projected to reach about 30 percent) largely because the career is extraordinarily demanding — long hours, unpredictable emergency call, and a mental absence even when physically home — that conflicts with the freedom many women want for childbearing and raising children, though many women neurosurgeons do successfully combine the career with motherhood.

    “They’re now at maybe 10 percent. I think we expect them to be at about 30 percent in a few years.”