Science & Tools of Learning & Memory | Dr. David Eagleman

The episode opens not with a definition but with a corrective. Asked whether doing crossword puzzles is good for an older brain, Eagleman answers that "it's good until you get good at it, and then stop and do something that you're not good at," and urges listeners to "constantly find the next thing that's a real challenge for you." The exchange, repeated later in the conversation, fixes the organizing claim of this stretch of the interview: plasticity is driven by novelty and difficulty, not by the comfortable repetition of a mastered skill.

Huberman's introduction frames why the mechanism matters. He names neuroplasticity as "your brain's ability to change in response to experience or any form of deliberate learning," and says that understanding the mechanisms behind it will "position you to make better decisions" and allow listeners, "if you choose, to rewire your brain to be a more effective learner." He also advertises the ground still to come — memory formation, stress and time perception, slow-motion perception in traumatic events, polarization, false memories, deja vu and dreams — so this first movement of the conversation is deliberately about the learning machinery itself.

Eagleman grounds plasticity in an evolutionary claim. He calls it "Mother Nature's big trick with humans," the decision to "drop a creature into the world with a half-baked brain and then let the world wire up the rest of it." He pairs this with the 1953 scene of Crick and Watson at the Eagle and Child pub claiming "We've discovered the secret to life" over the structure of DNA, then argues that DNA was only "half the secret of life," because "the other half is all around us" — culture, language, neighborhood. He notes in passing that he worked with Crick at the Salk.

His thought experiment makes the point concrete: born 30,000 years ago with exactly the same DNA, "would you be you? The answer is you wouldn't be," because a different culture, language and set of stories would make "a very different kind of person." The contrast animal is the alligator, which "born 30,000 years ago, you'd be the same alligator." Human flexibility, in his account, is why "we are the dominant species" and why each generation can absorb prior discoveries and "springboard off of that.

At the cellular level he offers the scale rather than a diagram: the brain has "86 billion neurons," each "on average contacting 10,000 of its neighbors," and these are "plugging and unplugging and searching around and finding new places to plug in," also changing the strength of their connections. The brain is "constantly reconfiguring itself every second of your life" — a picture of "little creatures" slithering in the head, which he offers as the mechanism by which "we absorb every single thing in our worlds."

A small disagreement opens the discussion of cortical hardware. Huberman reports the common framing that the human brain is "kind of like a macaque monkey brain with a supercomputer added on top of it," mostly prefrontal cortex. Eagleman corrects him: "It's actually cortex in general," and Huberman concedes, "Not just prefrontal cortex." The claim is quantitative: humans have "four times as much cortex as our nearest neighbors in the animal kingdom," and the cortex is only "the outer three millimeters of the brain," the wrinkly surface.

The deeper argument is that cortex is uniform. Eagleman calls it "a one-trick pony": it "looks the same everywhere is because it is the same," with "the same circuitry," "six little layers," and "the same algorithms," and "it gets defined by what you plug into it." Feed it visual input and it becomes visual cortex; feed it auditory input and it becomes auditory cortex. The textbook labels — visual, auditory, somatosensory — therefore describe inputs, not fixed hardware.

His key evidence is a rewiring experiment: "this study in 2000 by Mriganka Sur at MIT," in which, in a ferret, "he ... took the visual information, the optic nerve, and he plugged it into the auditory cortex," after which "what would have been the auditory cortex became visually responsive, and it started caring about vision." He draws the general conclusion that "the cortex is a one-trick pony," and says humans simply have much more of it, including prefrontal cortex.

He draws two consequences from that surplus of cortical "computational real estate." First, more room "in between input and output": where a squirrel, cat or macaque has sensory cortex next to motor cortex and simply eats the food placed before it, a human can say "I'm on a diet, I'm trying ... I'll eat it later." Second, the prefrontal cortex lets us "simulate what ifs" — to "think about possible futures, simulate things in a way that we don't have to risk our lives doing it," judging "that would be a bad idea" or "that'd be a pretty good idea" before acting.

Huberman tests the claim against a human case and Eagleman confirms it. Huberman notes that while cortical areas are "genetically devoted by virtue of wiring when we arrive in the world," there is "a lot of crossover, especially in the extreme cases": in someone blind from birth, the cortical real estate for vision "becomes allocated to tactile sensation, especially if they learn how to Braille read, maybe auditory processing." His conclusion — "there's really no blank real estate in the cortex. It's all used." — is endorsed as "exactly right."

Eagleman extends the takeover evidence. In people born blind, "what we call the visual cortex at the back of the head here ... gets taken over," becoming "devoted to hearing, to touch, to memory," with measurably finer discrimination; in people who go deaf, "the auditory cortex, all that real estate, nothing lies fallow in the brain," and can support feats such as reading an accent from lip movements. He offers savantism in autism as "one hypothesis" about the same principle: devoting "a ton of real estate" to a narrow skill such as the Rubik's Cube, piano or memorizing visual scenes can yield "absolutely superhuman" performance "at the cost of other things, let's say social skills."

Read together, the concrete evidence invoked in this section is a set of demonstrations that cortical function follows input rather than being fixed in advance. Eagleman does not present these as quantitative estimates, and he explicitly labels the savant explanation "one hypothesis."

The evidence cited, in the interview's own terms: — the 2000 Mriganka Sur ferret study at MIT (optic nerve redirected to auditory cortex, which became visually responsive); — congenital blindness, with Braille reading and occipital real estate devoted to touch, hearing and memory; — congenital deafness, including lip-reading of regional accents; — savantism in autism as a hypothesis about concentrated cortical allocation, carrying a trade-off in social skills.

Huberman raises early specialization as a cultural assumption worth probing — Tiger Woods, the Williams sisters, "especially in the United States, we have this notion that early specialization is really what sets you up to be spectacularly good later" — and asks whether genetic leanings make some activities "right for us or more right for us." He recasts the issue as a "funnel": early experience determines "the kind of width of the funnel" of later opportunity, depending on "how many things we did or didn't do early on."

Eagleman largely affirms the practice claim. The Williams sisters "got drilled on tennis from day one, and this stuff can be taught," the same pattern he sees "with chess champions and golf champions like Woods." Cognitively you can understand "what a forehand or a backhand is," but "to actually get good at it, you have to burn it down into the circuitry" — "you're taking that from the software to the hardware of the brain."

He explains that transfer through an efficiency argument. Playing Serena Williams, "she's beating me like crazy, but my brain's the one using all the activity. I'm the one burning all the calories," because "she has burned tennis into the hardware of the brain, so it's fast and efficient," while the amateur is "trying to simulate lots of things." Plasticity exists, in this account, because the brain's "main job is to save energy because we are mobile creatures who run on batteries," so repeated practice makes actions "fast and efficient."

He cites the Polgar sisters, three chess champions whom he calls, with a hedge, "to my knowledge ... still the best three female chess players in the world," taught chess by their father "from day one." On diversification he is more equivocal: he "can see why it would be useful" by analogy to learning both snowboard and ski, "you might get better at both of them," but he then supplies a limiting finding — children who grow up "trilingually, or even bilingually" "end up having a lower vocabulary in both languages than if they grow up monolingually."

That trade-off is attributed to practice time, not innate capacity: "It's just because of the amount of practice you get with a language," to which Huberman's wry response is "Kids, still do your second language homework." The surrounding exchange treats a second language and a musical instrument as worthwhile precisely because they force effort: Huberman says the "struggle of it ... is enriching in a way, because you're forcing your brain to do some work," and he quips, against students who cite "Google Translate or my Meta sunglasses," that "Google Translate is not Google Relate."

A personal anecdote introduces reward. Eagleman's father "spoke eight languages fluently without accent" because medical school in Europe and clinical rotations in different countries gave him, as a young man, "the incentive to learn the language" wherever he "got a girlfriend." Eagleman uses this to flag that "the reward systems are a big part of what makes change happen in the brain," a thread he explicitly sets aside as tangential before returning to it.

He then takes a contrarian position on the internet and education. Against people who have, "for the last 30 years, ever since the internet became a big thing," worried about what it means for children, he is "very optimistic": children gained "this opportunity to learn about something right when they were curious about it," whether fixing a bicycle tire or a question about space physics. The mechanism he proposes is that "brain plasticity really happens when you have the right cocktail of neurotransmitters present," and "that cocktail happens to map onto curiosity or engagement."

He contrasts this with the old school model in which a teacher "just dump[s] everything" — "the Battle of Hastings happened in 1066" — a fact "you may or may not ever need to know." Information delivered "right in the context of their curiosity" is what "makes a big difference because stuff really sticks." He reports being "extraordinarily impressed with young people" who know things via "TED Talks," "asked Alexa," or "talked to ChatGPT," concluding "and they get the information, and it sticks."

Huberman offers a partial counterweight from his own schooling: he remembers having to "bike or skateboard down to Tower Books" or "go to the library and look things up," and tells himself "the effort involved in going to get it actually is useful," while conceding that instant access would probably have left "more time implementing the information." Eagleman adds a social consequence: dinner-table arguments used to stall at "no one knows the right answer," whereas "now everyone whips out their phone, gets the answer, and then it keeps going," dissolving "some of the social dominance" that came from whoever "say[s] it with more certainty."

Huberman asks what he calls "the core 10 boxes of neuroplasticity" for building "a healthy, well-rounded nervous system," acknowledging "This is a tough question, I realize." The pair first settle the boundary conditions: infants "have no control over" their inputs, while adolescents, 20-year-olds and adults have "some control," though "assuming plasticity extends into adulthood, still as adults, although it's harder."

Eagleman gives two answers. The first is to "try to maximize along every axis" — "try to be an athlete, try to be a scholar, try to be somebody who's good at social life" — because "there are a million ways to waste time," illustrated by his airplane seatmate "playing Candy Crush for the whole flight" when "you could be reading books, you could be listening to podcasts." The second is forward-looking: since "who the heck knows what careers are going to exist in 20 or 30 years from now," the durable targets are "critical thinking and creativity."

For critical thinking he proposes "AI debate" as a route to "really individualized education," motivated by the observation that "In any classroom, it's going too fast for half the kids and too slow for the other half of the kids." A student would debate "any hot-button issue, abortion, gun control, whatever," be "graded based on the quality of your arguments," then "switch sides, and you take the other side." The stated aim is "so that students get a 360 view of issues instead of ideological capture," and he notes this is "the kind of thing you could never have enough teachers for."