Theodore Schwartz on Neurosurgery, Consciousness, and Brain-Computer Interfaces

Theodore Schwartz — neurosurgeon and professor at Weill Cornell Medical Center / New York-Presbyterian Hospital, author of Gray Matters: A Biography of Brain Surgery — joins Tyler Cowen in Ep. 243 to discuss the craft and psychology of operating on the brain, why surgeons resist outcome measurement, what epilepsy reveals about religious experience and free will, and how close we are to commercially viable brain-computer interfaces.

Key ideas

1. Surgical skill varies enormously, and medicine has almost no mechanism to measure it. Once the operating-room door closes, there is no systematic tracking of what a surgeon actually achieves — whether a tumour is fully removed, whether function is preserved. Hospitals measure what governments reimburse: infection rates and readmissions. The decisions that matter most — how aggressive to be, how much risk to accept — are invisible to every external observer, and reimbursement is flat regardless of experience or outcomes.

2. The long training pipeline is not merely hoop-jumping, but it could be compressed. Medical school plus neurosurgical residency can run to 16 years. Schwartz concedes undergraduate and medical school could be condensed to six years without loss; he is less certain about surgical training, arguing that maturity and judgement — not just technique — take time to develop, and that you want your surgeon to have lived enough to understand what losing a patient means.

3. The unified self is a story the brain tells after the fact. Schwartz draws on split-brain surgery and hemispherectomy to argue that the brain is not a single decision-making entity but hundreds of modules operating in parallel. What we experience as selfhood and free will is a narrative assembled post-hoc by one module to make the others’ outputs feel coherent. Consciousness — why there is something it is like to be a particular person — remains, for him, genuinely unexplained.

4. Deep brain stimulation can alter personality and desire, not just movement. Electrodes in the right circuits have suppressed and restored musical preferences, compulsive gambling, and obsessive behaviours. Any aspect of personality, Schwartz argues, is in principle alterable by changing the circuitry it runs on — a direct implication of a thoroughgoing brain-as-information-machine view.

5. Brain-computer interfaces are near-term for paralysed patients, far-term for enhancement. The core technology — recording from cortical neurons to drive robotic limbs or synthesise speech — is already demonstrated in academic settings. Neuralink’s advance is flexible electrodes that move with the brain and avoid scar formation. Schwartz estimates 15–20 years before elective neural interfaces are safe enough to offer to healthy people, but sees no principled reason to regulate them differently from mobile phones once safety is established.

Content

The training pipeline and selection

Medical school runs four years for everyone; neurosurgical residency adds six to eight more, typically seven, with one or two years of research folded in. Schwartz was 33 before he finished. He agrees with Cowen that the undergraduate and medical-school years could be condensed to perhaps six years in total without material loss — his philosophy and English degree was wonderful but not necessary for surgery. He is less willing to shorten residency itself: surgery requires not just technique but judgement, and judgement, he believes, requires a certain amount of lived experience.

The selection process at top programmes weights research ability heavily. Schwartz thinks this is incomplete. Neurosurgery is a physical, tactile sport — it demands stamina, coordination, grit, and the capacity to function well at 6 am and 9 pm alike. He would add physical and psychological endurance assessments to the 20-minute interview that currently dominates resident selection. IQ, he notes, is not the differentiator: a British Medical Journal study found neurosurgeons, rocket scientists, and average professionals all clustered at similar scores. What sets great surgeons apart is focus, dedication, and the ability to operate under high stakes without flinching.

The opacity of surgical quality

No institution systematically tracks what neurosurgeons actually accomplish in the operating room. Hospitals measure what governments pay them to measure: infection rates, readmissions, leg clots. Whether a surgeon removes a tumour in full or leaves tissue behind to protect motor function — a constant push-pull decision Schwartz describes as a ‘Sophie’s choice’ — is untracked. Reimbursement is flat across experience levels: a newly qualified surgeon and a 30-year veteran receive the same fee for the same procedure.

Schwartz believes outcome data could be collected and would be valuable; he does not believe the current system is remotely close to collecting it. He suggests the most reliable route to a good neurosurgeon is via someone in the medical profession who can call around — neurosurgeons know who the top operators are for any given procedure, because they are the people giving the most lectures and publishing the most articles. Google returns oddball results.

Consciousness, free will, and the storyteller brain

Asked whether the split-brain phenomenon means we are really two people, Schwartz reaches a more general conclusion: there is no unified self in the sense most people assume. The corpus callosotomy — the surgery that severs the two hemispheres — produces a patient whose left and right hands are controlled by different brains, yet who still reports feeling like one person. Hemispherectomy removes half the brain entirely; the patient does not wake up feeling like half a person. The self, Schwartz concludes, is a story assembled by one particular module — a ‘storyteller’ — to make sense of decisions that have already been made elsewhere.

He does not think the storyteller module has free will either. Ideas and decisions, in his view, emerge into consciousness from processes we do not understand and did not initiate. Consciousness itself — why there is something it is like to be any particular brain — he regards as genuinely mysterious and unexplained, not merely technically unsolved.

This view is grounded not only in surgery but in his research on epilepsy. Dostoevsky, a documented epileptic, experienced certainty of God’s existence during seizures; when his epilepsy was treated, the religious experiences disappeared and then recurred when the underlying lesion returned. Schwartz is not dismissing religious experience per se, but the case illustrates how directly brain state determines what we believe and feel.

Deep brain stimulation and its implications

Parkinson’s disease is already treated by inserting an electrode into the brain to suppress tremor. The same technique is being extended to depression, obsessive-compulsive disorder, Tourette syndrome, and addiction. Schwartz describes one documented case where switching a stimulator on made a patient suddenly prefer a particular musician; switching it off removed the preference entirely.

If desire can be created and extinguished by an electrode in the right place, then, he argues, any aspect of personality is in principle alterable: all personality runs on brain circuitry, and circuitry can be changed. He also discusses focused ultrasound — a technique that opens the blood-brain barrier in targeted anatomical regions — as a near-future route to delivering drugs specifically to the depression-relevant circuits rather than diffusing them through the whole brain.

Electroconvulsive therapy (ECT), he notes, works on similar logic: a whole-brain seizure resets network functioning in ways that alleviate severe depression. It is crude, and he expects it to be superseded, but it is currently one of the most effective tools available for refractory cases.

Brain-computer interfaces and Neuralink

Schwartz distinguishes two classes of BCI device: those that insert electrodes into the brain parenchyma (Neuralink’s approach, recording from individual neurons) and those that lay electrodes on the cortical surface. Both capture useful signal; neither is yet definitively superior. Neuralink’s specific technical advance is flexible electrodes that move with the brain’s pulsation, avoiding the dense scar tissue that older rigid electrodes provoked.

The technology as a research tool is further along than most people realise. Academic groups have already demonstrated: paralysed patients driving robotic arms; patients with locked-in syndrome generating synthesised speech from neural activity in real time. The remaining challenge is miniaturisation and wireless transmission — removing the wires that currently protrude from the skull.

For healthy people seeking cognitive enhancement, Schwartz is more measured. He estimates 15–20 years to make the surgery safe enough to offer electively. The regulatory question is, to him, ultimately similar to the mobile phone: once safety is established, he sees no principled reason to require government permission to buy one. The harms that might warrant regulation are the downstream uses — what you can do with the interface — not the interface itself.

The philosophy of mind and Freud

Schwartz grew up in a household shaped by a Freudian psychoanalyst father. He reads widely in philosophy of mind and follows contemporary debates on consciousness, self, and agency. On Freud circa 2025: he thinks Freud is underrated if you focus on the structural insight — that unconscious processing dominates our behaviour and the ego is a small, rationalising layer on top — rather than on the feminist critiques of penis envy and the Oedipal complex. Neuroscience, he argues, has confirmed Freud’s core intuition that a great deal of what we do is decided below the level of awareness.

He closes by noting he is starting a medical device company, hoping to reach not one patient at a time but hundreds of thousands.

Related

• Theodore Schwartz — speaker; neurosurgeon at Weill Cornell Medical Center, author of Gray Matters
• Tyler Cowen — host
• Consciousness — concept; the hard problem of consciousness, why there is something it is like to be a person, discussed directly
• Free Will — concept; Schwartz’s view that decisions are made by unconscious modules and narrated post-hoc