Why AI's builders want medicine first: restoring lost function before augmentation

The Short Version

The people actually building devices that plug into the nervous system all start with patients who have lost something. Neuralink's head neurosurgeon, Matthew MacDougall, describes a first product meant to let quadriplegic people control a computer with their thoughts [1]. Stanford's E.J. Chichilnisky is building a "smart retinal implant to restore sight to people whose photoreceptors have died [2]. Casey Halpern at the University of Pennsylvania uses deep brain stimulation to interrupt the loss-of-control urges behind binge eating, OCD and addiction [3]. Each of them talks about enhancement, if at all, only after the medical work, and even Mark Zuckerberg, whose company sells consumer hardware, expects brain implants to get started in medical use cases long before anyone is jacking people into play games better [4]. The shared reasons are that the risk is only worth it when the benefit is large, that regulators require safety proof first, and that restoring a lost function is how scientists learn enough about a circuit to change it. Taken together, the sources suggest brain-machine interfaces will reach the public through the clinic, and that consumer upgrades" will either follow later or come through non-invasive devices worn on the body [1][2][4].

Table of contents

What They Said

Matthew MacDougall (Neuralink): restore "digital freedom" first, dream later

MacDougall is Neuralink's head neurosurgeon, trained at UC San Diego and Stanford, and he spends much of his conversation with Andrew Huberman taking the mystique off the company [1]. He says Neuralink, like Tesla and SpaceX, becomes a blank canvas for people's hopes and fears, and that both extremes are mostly wrong [1]. What the company is actually building, he explains, is a neural implant plus a robot that inserts electrode threads thinner than a human hair into a small region of the brain [1].

The first patients. The first target is people with severe spinal cord injuries who are quadriplegic: their motor cortex is intact but disconnected from their limbs [1]. The implant would read their movement intentions so they can control a computer mouse and keyboard, regaining what MacDougall calls digital freedom [1]. He is careful about the limits of that first step: at this stage patients will not move their own limbs; imagining a movement moves a cursor, opens a word processor and types on an on-screen keyboard [1]. He describes the company's broad near-term mission as reducing human suffering, with using AI as a tool embedded in the brain as a longer-term hope [1].

The ladder of goals. MacDougall lays out his ambitions in order. The next medical step is to reconnect the brain to the body by pairing the brain implant with a spinal cord implant below the injury, and Neuralink has done considerable work on spinal cord stimulation toward this [1]. Over 10 to 20 years, he wants to give people control over the ways the brain goes wrong, such as addiction, depression, suicide and obesity [1]. Only for his children's lifetimes does he imagine full expansion of human cognition into AI, instant access to information without reading up first, communication freed from what he calls the narrow straw of speech, and perhaps even opt-in linked minds [1]. Brain-to-brain messaging, he says, is mostly an engineering problem, since two people texting each other already communicate silently; what matters is bit rate, how much useful information flows in and out [1].

Why the order is forced. Several practical constraints push the work toward medicine. Current technology can reach only the brain's surface, so deep functions like mood, appetite, addiction, pain and sleep will come later [1]. The FDA requires proof of safety in animals first; MacDougall says the company is full of obsessive animal lovers and would avoid animal research if there were any other way [1]. Pigs are used mainly as a safety platform because their skulls are similar in size and shape to humans', while monkeys provide the functional data [1]. And trials in people begin with patients: he mentions a patient registry for people with quadriplegia interested in future trials, which were still being finalized with the FDA at the time [1].

Would he use it himself? When Huberman asks whether he would implant himself, MacDougall says absolutely, but that it would not be useful yet because early devices do things he can already do with his hands [1]. Once a device lets you type faster than speech-to-text or your fingers, it would make sense for someone like him [1]. This is the clearest statement in the sources of the tipping point between restoration and enhancement: the device becomes attractive to a healthy person only when it beats the body's own performance [1]. He adds that, from the hundreds of surgeries he has performed with it, the procedure is already safer than many standard FDA-approved operations [1].

The clinic as teacher. MacDougall's view of the brain also comes from patients who lost things. He describes it as a collection of functional modules, which becomes obvious when a small tumor knocks out one very specific function [1]. Early in his training he helped save a young man after a car accident who woke with damage to both frontal lobes and no remaining impulse control; MacDougall says it was clear the man would never behave normally again [1]. He calls the frontal lobes a filter that tells the rest of the brain to hush, letting impulses through only in a controlled, appropriate way [1]. He also gives a striking example of a deep brain function gone wrong: children with hypothalamic tumors can suffer gelastic seizures, uncontrollable humorless laughter so severe that some have passed out or suffocated [1]. Every one of these is a function a future device might one day restore, and every one was learned from loss.

A surprising limit on enhancement. When Huberman asks about boosting plasticity in adults, MacDougall gives what the summary calls a surprising answer for a Neuralink surgeon: he thinks drugs, including the psychedelics Huberman mentions, will be more fruitful than implants [1]. Changing plasticity means altering something like a trillion synapses across the whole brain, which a drug can reach and an electrode cannot, and he does not think even steerable electrodes will cover the brain that way within our lifetimes [1]. Neuralink is not currently combining drugs with stimulation, because decoding the brain electrically is hard enough [1].

E.J. Chichilnisky (Stanford): understand the retina so well you can replace it

Chichilnisky, a Stanford professor of neurosurgery, ophthalmology and neuroscience, studies how the retina encodes the visual world and uses that knowledge to build artificial retinas that could restore sight to the blind [2].

Why start where function is lost. A major cause of blindness in the Western world, including macular degeneration and retinitis pigmentosa, is the death of photoreceptors [2]. The long-standing idea, which he says is not his invention, is to bypass the damaged early layers with an implant that captures images with a camera, processes them the way the retina would, and electrically stimulates the ganglion cells to send signals to the brain [2]. Asked why he focuses on the retina rather than the visual cortex or thalamus, he says: "I focus on the retina because I enjoy the possibility that we can really understand a piece of the nervous system in my lifetime, in our lifetimes. We can understand it so well that we can build it, replace it, restore its function." [2] He adds that the goal is to understand it so well "that I can engineer devices to replace the function of that circuit once lost." [2]

The honest state of the art. Pioneering engineers have already made retinal implants work in patients: people blind for decades can see blobs and streaks of light, enough to find a bright doorway or avoid a coffee table [2]. That is the glass half full [2]. The glass half empty is that this is nothing like natural vision, because existing devices treat the retina like a camera's grid of pixels and stimulate whatever cells happen to be near each electrode [2]. He compares the retina's output to an orchestra in which each cell type has its own score, and says current implants are as if the conductor had scattered the sheet music everywhere [2]. Asked how decades of publicly funded retinal science show up in vision-restoring devices, he answers: "Well, currently it's not. And so we're trying to do that." [2] He believes scientists have a duty to give back by turning that knowledge into something that helps people [2].

The smart implant. His goal for the coming decade is a device that respects the retina's circuitry, working in three steps: record which cells and cell types are present in that particular person with a chip roughly two millimeters in size; calibrate by building a table of how each electrode activates each cell; then stimulate each cell in the right sequence when an image comes in [2]. The device, he says, must "recognize the distinct cell types. See where they are. Stimulate them separately from one another and conduct this orchestra to create a musical score that reasonably closely resembles the natural one." [2] His slogan is "Make our instruments smart," rather than expecting the brain to wrap itself around a dumb device [2]. He and Stanford engineers have spent about 15 years on the basic science and are now working to build a device for living humans [2].

Augmentation comes second, and uninvited. Chichilnisky is the source who talks most openly about enhancement, and he places it explicitly as a spin-off of restoration. The mission of restoring the retina's normal patterns, he says, "also has extremely exciting spin-offs in three directions" [2]: understanding how the brain puts signals together; "augmenting vision, creating novel types of visual sensations that weren't possible before and maybe doing something along the Elon Musk lines of delivering more visual information than was ever possible"; and interfacing with the brain more broadly [2]. His thought experiment: because texting while driving is dangerous when both tasks compete for vision, a device might one day send text to one pathway (the tiny "midget cells) and moving traffic to another (the parasol" cells), streaming parallel information into the visual system [2]. He stresses this is not a research goal for texting, only an illustration [2].

He also argues there is no bright line between restoring and augmenting. The National Institutes of Health generally wants to restore rather than create a new kind of human, yet even the crudest vision-restoration device must filter out infrared light, since many cameras detect it, or users would gain infrared vision [2]. In his words: "as soon as you start building devices to restore sensations, building electronic devices, augmentation is right around the corner. It'll creep up on your real fast. So I think you can't even really draw a line." [2] He sums it up as It's not sci-fi anymore. It's sci, and says that because humanity will clearly build such technology, scientists have a responsibility to do it thoughtfully [2].

Slow and gradual. Even his vision of augmentation is incremental. Drawing on Eric Knudsen's finding that adult animals show plasticity when sensory changes are introduced gradually, he expects a new visual signal would need to be dialed up slowly [2]. Once there is full control of the retina's code, he says, "we can then start to play games and dial up that neural code very slowly and teach the remaining brain how to understand these signals, not introduce some crazy thing from day one." [2] Huberman contrasts this grounded approach with hype about chips for super memory [2].

Andrew Huberman: the retina as test bed, and a spectrum of specificity

Huberman, who hosts two of the conversations, adds his own reasoning for why medicine leads. He compares the retina to the hippocampus, where people fantasize about stimulating a chip for super memory: there is decent knowledge of hippocampal cell types, but nothing close to the understanding of what each does, so stimulating it would produce unpredictable effects [2]. The retina, by contrast, is the perfect test bed for neural engineering, and it is physically accessible at the body's surface [2]. As he likes to remind listeners, "when you're looking at somebody's eyes, you are looking at two pieces of their brain." [2]

He also frames a spectrum of brain interventions by specificity: drugs that raise serotonin or dopamine act on receptors throughout the brain, which is why they bring side effects along with benefits; electroconvulsive therapy, still used for depression and sometimes effective, is even less specific [2]. Chichilnisky likens shock therapy to rebooting a malfunctioning computer: it might work, but you would rather diagnose the software and fix the specific problem [2]. Cell-type-specific stimulation sits at the other extreme [2]. Closing that conversation, Huberman describes the arc as moving from curiosity to "a mission in service to humanity, restore vision to the blind, develop neuroprostacies and other types of neuroengineering technologies that will allow the human brain to function better than it would otherwise." [2] Note the order in his own sentence: restore first, then function better.

Casey Halpern (Penn): deep brain stimulation for lost impulse control

Halpern, Chief of Neurosurgery at the University of Pennsylvania School of Medicine, works on compulsive eating, binge eating disorder, bulimia and other obsessive-compulsive behaviors [3]. His starting point is patients who have lost control. He explains that compulsive eating is not simply a matter of willpower but involves specific neural circuits that create irresistible urges to eat, even when a person knows they shouldn't or doesn't want to [3]. The key areas are the nucleus accumbens, the brain's reward center; the prefrontal cortex, responsible for impulse control and decision-making; and the basal ganglia, which regulate habit formation and compulsive actions [3]. In binge eating disorder, the nucleus accumbens and prefrontal cortex fail to regulate these urges [3].

His research led to the first-ever human trial using deep brain stimulation (DBS) to treat binge eating disorder and compulsive eating [3]. Electrodes are placed in the nucleus accumbens; the device detects neural activity associated with cravings and delivers a brief, targeted pulse of electrical stimulation when an urge is detected, disrupting the craving circuit [3]. In animal models, DBS blocked binge eating behaviors, and in human trials patients reported a significant reduction in binge urges and increased control over food intake, according to the summary [3]. Halpern stresses that binge eating, OCD and addiction share one feature, the inability to stop an urge despite negative consequences, and that these conditions exist on a spectrum, so treatments that work for one may work for others [3].

Because DBS requires surgery, researchers are also exploring non-invasive options: transcranial magnetic stimulation, described as FDA-approved for depression, OCD and nicotine addiction and being tested for compulsive eating and obesity; focused ultrasound, described as FDA-approved for essential tremor and Parkinson's disease and being explored for OCD, binge eating and addiction; and wearable devices that might one day use AI to detect early craving signals and deliver stimulation in real time [3]. Unlike the other sources, the summary of Halpern's conversation contains no discussion of enhancement at all: the entire frame is treatment [3].

Mark Zuckerberg (Meta): implants start with patients, wristbands go to consumers

Zuckerberg is the only speaker here who runs a consumer hardware company, and he gives the most explicit version of the medical-first prediction. Asked by Joe Rogan whether a neural interface could be the bridge to fully immersive experiences, he says there are two approaches, roughly: "the kind of jacket into your brain neural interface and there's the wrist-based neural interface thing" [4]. On the first, he says: "I think it's going to be a while before we're really widely deploying anything that jacks into your brain. I think that there are a lot of people who don't want to be the early adopters of that technology." [4] He continues: "that's basically going to get started in medical use cases, right? So if someone like loses sensation, part of their body and now you have the ability to fix that, like the first neural-length patient." [4] And he states the logic directly: "you'll basically start with people who have pretty severe conditions, who the upside is very significant. Before you start, like, jacking people into play games better, right?" [4]

His company's consumer bet is therefore something else: a wristband that reads the signals the brain sends to the hand [4]. He explains that the hand has many extra neural pathways because of redundancy, "because being able to use your hand is super important," and that a band on the wrist can measure that activity [4]. Today it measures signals as you move your fingers, but over a few versions, "you won't actually even have to move your hand. You'll just like trigger these neurons in opposing ways." [4] The appeal is that "people wear stuff on their wrist all the time," and he is clear about its limits: "that doesn't do input to you, but it's good for giving you the ability to control a computer." [4] He envisions discreetly texting or querying AI and seeing answers in glasses [4]. Rogan, who tried the Orion glasses and neural band, describes playing pong across a table with Lex Fridman [4].

Where They Agree, and Where They Don't

Agreement: start with loss, with the biggest upside

All four sources begin with a patient population defined by a lost function: movement and computer access for quadriplegia [1], sight for photoreceptor loss [2], control over urges for binge eating, OCD and addiction [3], and, in Zuckerberg's example, someone who "loses sensation" [4]. Zuckerberg supplies the plainest reason, the risk-benefit trade: start with people "who the upside is very significant" [4]. MacDougall supplies the regulatory one: the FDA requires animal safety data before any human work, and trials run through patient registries [1]. Chichilnisky supplies the scientific one: restoring a circuit you understand is the only honest route to changing it, and even the NIH's funding frame is to restore rather than create a new kind of human [2].

They also agree that enhancement is real but later. MacDougall's AI-augmented thought is a hope for his children's lifetimes [1]; Chichilnisky's augmented vision is a "spin-off" of restoration that must be dialed up slowly [2]; Zuckerberg puts gaming implants after medical ones [4]. And MacDougall and Chichilnisky both place the work on a long academic foundation: MacDougall and Huberman credit a century of basic research, including Apostolos Georgopoulos's early recordings from motor cortex and the late Stanford researcher Krishna Shenoy [1], while Chichilnisky independently cites Shenoy and Jaimie Henderson at Stanford, Eddie Chang, and Neuralink building on that research for letting paralyzed people move cursors or communicate [2]. As Chichilnisky puts it about those patients: "with their thoughts can send signals through an electronic device that can be used to control a mouse on a screen and have them connect to the internet." [2]

Disagreement 1: is there a line between restoring and enhancing?

Here the sources genuinely differ. Zuckerberg and the NIH framing described by Chichilnisky treat the two as distinct stages: medical first, gaming or new kinds of humans later [2][4]. Chichilnisky himself argues "you can't even really draw a line," because any device that restores a sense must make choices, such as filtering infrared, that could just as easily add something new [2]. MacDougall sits in between: he sees a clear practical threshold (the device must beat his fingers before he would want it) but no principled one, since he would take it the moment it did [1]. The difference comes from where each stands. Chichilnisky is designing the signal itself, so he sees that every design decision shapes perception; MacDougall and Zuckerberg are thinking about who will accept surgery, which is a question of benefit versus risk.

Disagreement 2: implants versus drugs, wristbands and non-invasive tools

The sources also disagree about whether electrodes in the brain are the best tool at all. MacDougall, remarkably, says drugs will be more fruitful than implants for boosting plasticity, because a drug reaches a trillion synapses and an electrode cannot [1]. Chichilnisky and Huberman argue the opposite direction on a different question: drugs and electroconvulsive therapy are blunt, and cell-type-specific stimulation is the precise end of the spectrum [2]. These views are not strictly contradictory, since MacDougall is talking about broad changes in plasticity and Chichilnisky about precise control of one circuit, but they point to different futures for "enhancement."

On consumer adoption, Zuckerberg bets on the wrist rather than the skull [4], and Halpern's field is already reaching for non-invasive options such as magnetic stimulation, focused ultrasound and wearables precisely because DBS requires surgery [3]. MacDougall answers this camp directly: when Huberman mentions Neosensory, a wrist device that translates sound into touch and helped a deaf woman feel birdsong, he says such peripheral devices have lower barriers to adoption but low ceilings on bit rate, while implants could potentially interface with millions of neurons [1]. So the likely split, reading all sources together, is that the high-bandwidth implant path runs through medicine, while the consumer path runs through lower-bandwidth wearables that skip surgery [1][3][4].

An exception worth noting

MacDougall himself is a small counterexample to the medical-first pattern: he implanted a passive RFID chip between his own thumb and index finger at his kitchen table, and it opens his front door and Neuralink's doors [1]. His wife asked for one too after seeing its convenience [1]. But the chip is passive, has no battery and sits in the hand, not the brain [1]. It shows that healthy people will adopt a low-risk implant for convenience; it does not show anyone accepting brain surgery for an upgrade.

Why It Works

Lost function leaves intact machinery to talk to

The medical-first path works partly because the target patients still have healthy tissue on one side of the break. In quadriplegia, the motor cortex is intact but disconnected from the limbs, so a device can read intentions that are still being generated [1]. In photoreceptor blindness, the ganglion cells that carry signals to the brain remain, so an implant can bypass the dead early layers and stimulate them [2]. In compulsive eating, the problem is an overactive reward circuit that fails to be regulated, so a device can detect the craving signal and interrupt it [3]. In each case the device fills a gap in an otherwise working system, which is simpler than adding a capability the brain never had.

The software and the brain learn each other

MacDougall explains that much of the difficulty is moved from the patient's brain into software that decodes motor intentions [1]. At first, firing patterns mean nothing to the chip, but by correlating them with what the person is asked to do in games (move the cursor up and right, down and left), the software builds a model the person then learns to use, even as signals drift from day to day [1]. He compares it to learning a piano whose keys shift over time, and Huberman likens it to biofeedback [1]. Human participants will speed progress because, unlike monkeys, they can tell researchers what they are trying and adjust [1]. Neuralink already has monkeys playing video games for smoothie rewards and, MacDougall says, holds the world record for the bit rate decoded from a monkey's brain to control a cursor [1]. He also notes that a Stanford group decoded speech from the hand area of motor cortex, showing each brain region carries many useful signals [1].

Precision needs deep understanding, which only some circuits have

Chichilnisky's mechanism is about knowledge. The retina has about 20 types of ganglion cells, each covering the whole scene and extracting different features: "You can sort of think of them as Photoshop filters," he says, so the brain receives 20 different versions of the world [2]. About seven types are well understood and carry perhaps 70% of the signals, while his lab member Alexandra Kling recently found about 15 more with "crazy properties" whose function remains unknown [2]. His lab builds this knowledge from donated human retinas, recorded over 48 nonstop hours on a 512-electrode array while random "TV snow is projected, and by passing current through the same electrodes in order to see if we can activate those ganglion cells directly with no light, just electrodes [2]. That is why he calls the retina the place where neuroscience can move from understanding to engineering, and why Huberman warns that a hippocampal memory chip" would produce unpredictable effects [2]. Chichilnisky also says AI helps with the transformation from image to neural code, but stresses that it is an engineering tool, not a source of understanding [2].

Surgical tools that make the risk acceptable

The engineering around the implant matters too. The electrodes are so tiny and the surface blood vessels so dense that no human hand is steady or fast enough to place them accurately at the right depth, so a robot does the placement while a neurosurgeon opens the skin and skull [1]. MacDougall also describes laser interstitial thermal therapy, which destroys deep tumors through a two-millimeter hole while an MRI monitors temperature in real time, so that tumors once considered inoperable can sometimes be removed with patients coming out essentially unscathed [1]. Less invasive surgery lowers the risk side of the ledger, which is exactly what, in Zuckerberg's framing, decides who goes first [1][4].

Closed-loop stimulation

Halpern's DBS system illustrates a second design principle: the device senses before it acts. It detects neural activity associated with cravings and delivers a brief pulse only when an urge is detected [3]. The summary of his conversation draws a parallel with self-awareness: recognizing early warning signs such as stress, anxiety or specific triggers can help people intervene before losing control, much as DBS works by disrupting craving signals [3].

Adult brains adapt, if changes are gradual

Finally, Chichilnisky explains why even restored signals, let alone new ones, must be introduced carefully. Eric Knudsen showed adult animals adapt when sensory changes come gradually, and spike-timing-dependent plasticity means the precise timing of neural activity shapes connections [2]. MacDougall adds that age and species matter enormously: babies can recover after losing half the brain, but adult humans are very vulnerable to losing discrete functions from small lesions [1]. Both points favor slow, clinically supervised introduction over sudden consumer upgrades.

Where It Applies, and Where It Doesn't

It applies to invasive, high-bandwidth devices. Brain surgery, implanted electrodes and deep stimulation are where every source puts patients first [1][2][3][4]. Zuckerberg's line that people don't want to be early adopters of anything that "jacks into your brain" applies squarely here [4].

It applies most to the brain's surface and best-understood circuits, for now. Current Neuralink technology reaches only the surface, so deep functions like mood, appetite, addiction, pain and sleep come later [1]. Halpern's DBS does reach deep, into the nucleus accumbens, but through a different, established surgical technique [3]. Chichilnisky's retina is favored precisely because it is accessible and well understood [2].

It applies less to non-invasive wearables. The wristband Zuckerberg describes, the Neosensory device MacDougall discusses, and the wearables Halpern's summary mentions all have lower barriers to adoption [1][3][4]. These may go straight to consumers, though MacDougall says their ceiling on bit rate is low [1], and Zuckerberg's band only controls a computer; it "doesn't do input to you" [4].

It applies less to trivial implants. MacDougall's hand-implanted RFID chip shows that healthy people will take on a small, passive implant for convenience [1]. That is not a brain-machine interface.

The line may blur sooner than expected. Chichilnisky warns that augmentation "It'll creep up on your real fast" with any restoration device, starting with choices like whether to pass infrared light [2]. And MacDougall says he would implant himself once a device let him type faster than his fingers [1]. The medical-first route is how the technology arrives, but it does not guarantee it stays medical.

Not every source addresses enhancement. The summary of Halpern's conversation is entirely about treatment and says nothing about augmenting healthy people [3], so it supports the "medical first" half of the pattern but offers no view on what comes after.

For anyone considering these treatments, these are specialist decisions: talk with a neurologist, neurosurgeon, ophthalmologist or psychiatrist about what is actually approved and available for your condition.

What To Do With It

1. Judge brain-interface news by who the patient is. When a company announces a device, ask what lost function it restores and for whom. MacDougall says Neuralink's real near-term product is computer control for quadriplegic people, and that both the hype and the fear around the company are mostly wrong [1].

2. Watch clinical trials and registries, not product launches, for the first real signals. Neuralink's early access runs through a patient registry for people with quadriplegia and trials negotiated with the FDA [1]; Halpern's DBS work began as the first-ever human trial for binge eating disorder [3].

3. If you or someone close to you has one of the target conditions, ask about trials. Quadriplegia [1], blindness from photoreceptor loss such as macular degeneration or retinitis pigmentosa [2], and treatment-resistant binge eating, OCD or addiction [3] are the conditions these sources name. A specialist can say whether a study fits.

4. Set realistic expectations for first-generation devices. Today's retinal implants give blobs and streaks of light, enough to find a doorway or avoid a coffee table, not natural vision [2]; Neuralink's first device moves a cursor, not limbs [1]. Improvements are coming, but gradually.

5. Try non-invasive options first where they exist. Halpern's summary lists magnetic stimulation and focused ultrasound as alternatives because DBS requires surgery, and highlights awareness of pre-craving states, cognitive behavioral therapy and mindfulness as tools for recognizing and resisting urges [3].

6. Expect consumer "neural" products to sit on the body, not in it. Zuckerberg's wristband reads signals to the hand and lets you control a computer [4]; MacDougall notes such devices are easier to adopt but carry less information [1]. Evaluate them as input devices, not brain upgrades.

7. Treat claims of memory or intelligence chips with suspicion. Huberman explains that stimulating the hippocampus would produce unpredictable effects because we don't know what each cell type does [2], and Chichilnisky insists any new signal must be dialed up "very slowly" [2]. If a pitch skips the restoration stage, it is skipping the science.

8. Protect the brain you have while you wait. The same surgeon building Neuralink names alcohol as far and away the most common brain damage people volunteer for, describing heavy drinkers whose scans show brains atrophied to look like small walnuts [1]. Huberman cites UK Biobank data linking regular drinking to gray-matter thinning from just one drink a week [1]. MacDougall also warns about focal blows to the thin temporal bone, "God's Little Joke" [1].

9. Support the basic science and donation systems that make restoration possible. Chichilnisky's lab depends on donated retinas and credits organ procurement organizations like Donor Network West [2], and both he and MacDougall stress decades of publicly funded academic research behind today's devices [1][2].

Sources

1. Dr Matthew Macdougall Neuralink Technologies To Enhance Human Brains · Huberman Lab
2. Dr. E.J. Chichilnisky: How the Brain Works, Curing Blindness & How to Navigate a Career Path · Huberman Lab
3. Dr. Casey Halpern: Biology & Treatments for Compulsive Eating & Behaviors · Huberman Lab
4. Mark Zuckerberg · The Joe Rogan Experience