how to improve memory: 6 tactics from the scientists who study it

Huberman opens with what he calls "One of the biggest lies in the universe": the idea that every experience changes your brain. after childhood, it does not. on the transcript, he says "if we want something to change, we really need to bring an immense amount of attention to whatever it is that we want to change." most forgetting is not forgetting. the thing never got in.

the tactic: train your eyes to train your attention. his rule: "mental focus follows visual focus." his own routine: "Spending just 60 to 120 seconds focusing my visual attention on a small window of my screen" before he works. and when your mind drifts, "attention drifts, but we have to re-enter it." you bring your eyes back to the page, again and again.

try this week: before your next reading or study session, look at one small spot on a blank screen or wall for 60 to 120 seconds. then start, and every time your eyes wander, bring them back to the line you are on.

Huberman asks why some things stick and others vanish. Suzuki's answer, on the transcript: "there are four things that make things memorable." "Number one is novelty." "Second is repetition." "Third is association." "And then the fourth one is emotional resonance." on that last one: "we remember the happiest and the saddest moments of our lives, and that also includes funny, surprising things."

the tactic: pull more than one lever on purpose. most of us only use repetition, reading the same thing over and over. the other three are free. make it new (a different place, a different format), tie it to something you already know, and give it a feeling: a joke, a story, a reason it matters to you.

try this week: pick one thing you need to remember this week, like a talk, a list or a chapter. write down how you will use all four levers on it: one new angle, one repeat, one link to something you know, one emotion. then check on sunday what stuck.

"I can't remember names" is the first complaint he hears. his take, on the transcript: "if you can't remember someone's name, that's probably a natural forgetting, but you remember if you had an interesting conversation, or you remember if you can trust them." names are hard because they are arbitrary. so you have to build a link: "Castel rhymes with Pastel, and my grandfather used pastels".

the tactic: build the link, then retrieve it. he admits the trick has a cost. it takes effort and can backfire (people call him Pastel). so pair it with practice: "we have to use deliberate practice, retrieval practice, use the name several times." and if a name still slips, remember what you talked about. he says that often matters more.

try this week: with the next three new people you meet, say their name back once, make one link (a rhyme, a person you know with the same name), and use it once more before you leave. that night, write down the names and one thing each person said.

you have seen the Apple logo thousands of times. Castel asks his students to draw it, and they struggle: which side is the bite on, is there a stem or a leaf? the struggle is the lesson. on the transcript, after a failed attempt, "when you look at the logo again, you're going to engage in better learning than if you hadn't done that error-full trial beforehand." and that holds even if you get the drawing wrong.

the tactic: fail first, then look. "this sort of retrieval failure with some corrective feedback is really beneficial." re-reading feels productive because the material looks familiar. but familiar is not the same as remembered. a blank page shows you the gaps, and the check right after fills them.

try this week: before you re-read any notes, a book chapter or a podcast summary, take a blank page and write down everything you remember. then open the source and mark what you missed or got wrong. that marked list is what you study.

in his focus episode, Huberman puts it bluntly: "neuroplasticity doesn't occur during wakefulness. It occurs during sleep." Poe explains when. the first stretch of the night carries a growth hormone pulse that you cannot catch up on by sleeping later. Huberman asks if that means consistent bedtimes. her answer, on the transcript: "Exactly. And in fact, one of the best markers of good neurological health when we get older is consistent bedtimes."

the tactic: protect the night your memories move. Poe describes sleep spindles as "really important for moving memories to our cortex." in her words, "our hippocampus, like the RAM of our brains, writes it to a hard disk which is the cortex." and she names one thing that gets in the way: "alcohol is a REM sleep suppressant." Huberman suggests none in the four to six hours before sleep. she agrees, though she is not sure of the exact dose response.

try this week: pick a bedtime and hold it within 30 minutes for seven nights, weekends included. no alcohol in the four to six hours before bed.

Suzuki's image, on the transcript: "every single time you move your body, it's like giving your brain this wonderful bubble bath of neurochemicals." her lab tested how little it takes. people who had been exercising under 30 minutes a week did "three months of two to three times a week cardio". their mood got better, and so did their scores on two memory tests. in a second study, people who added sessions up to seven a week kept improving: "every drop of sweat counted".

the tactic: treat cardio as memory training. Castel cites a study where one group walked three or four times a week for 40 minutes. "the walking group, their hippocampus actually increased in volume by 1%." and "Their memory was much better a year later." Suzuki adds that a single session gives you a focus boost that lasted up to two hours in her lab, so put hard learning right after a workout.

try this week: book two or three cardio sessions of 30 to 40 minutes (a brisk walk counts). put one before the hardest thing you need to learn or remember this week.