Getting it wrong is the signal: why errors, not flow, drive learning

The Short Version

The feeling that learning is going badly, the frustration, the missed attempts, the sense of being stuck, is usually the sign that it's working. Andrew Huberman explains that errors release the chemicals that tell the brain to change, so "errors are the basis for neuroplasticity and for learning", while smooth, correct performance gives the brain no reason to rewire [1][3]. Josh Waitzkin, the chess prodigy turned martial arts world champion, says his most devastating losses "were the greatest moments of my life in retrospect" because they seeded his later breakthroughs [6]. Growth mindset research shows the difference between people who improve and those who don't lies in how they treat mistakes: as something to analyze or something that stings [5]. The practical lesson for students and teachers is to read struggle as learning in progress, and to stay with it a little past the point where quitting feels natural.

Table of contents

What They Said

Huberman: errors are the signal to change

Across several episodes, Andrew Huberman, a Stanford professor of neurobiology, builds the case that the adult brain changes mainly when something goes wrong [1][2][3][4]. He frames his episode on failures, movement and balance around "why making errors is actually the signal that tells the brain, okay, it's time to change" [1].

His starting point is a question: "why would your nervous system ever change?" [1] It would have no reason to, he argues, unless something frightened us or made us feel awful, or unless there was an error in our performance [1]. The feedback from errors releases a specific chemical sequence [1]. Epinephrine increases alertness; acetylcholine creates "an opportunity to focus on the error margin, the distance between what it is that you're doing and what it is that you would like to do" [1]. The nervous system starts making changes almost immediately to get the behavior right, and "when you start getting it even a little bit right, that third molecule comes online or is released, which is dopamine, which allows for the plastic changes to occur very fast" [1].

This reframes frustration. "The brain doesn't even understand frustration as an emotional state. The brain understands the neurochemicals that are released namely epinephrine and acetylcholine," he says [1]. The conclusion follows: "errors are the basis for neuroplasticity and for learning" [1]. And it predicts who succeeds: "humans do not like this feeling of frustration and making errors. The few that do do exceedingly well in whatever pursuits they happen to be involved in. The ones that don't generally don't do well. They generally don't learn much." [1]

The choice at the moment of frustration matters. If you channel frustration into drilling deeper, you set up powerful plasticity; if you walk away frustrated, plasticity rewires you around whatever comes next, which is usually feeling miserable [1][2]. Staying with the process a bit past the point of frustration, he says, is the most important thing for adult learning [2].

Huberman: flow is not the learning state

In the longer version of that episode, Huberman goes further and separates learning from flow [2]. "Flow is an expression of what we already know how to do. It is not a state for learning." [2] The popular idea that we learn best when everything feels smooth gets it backward. Similarly, routine exercise maintains the nervous system but doesn't change it; behavior changes the brain only when it differs enough from what you already do well [2]. The way to create plasticity is to create mismatches, signals that something is wrong or not being achieved [2].

He makes the same point in conversation with Waitzkin. Most people associate being in flow with getting everything correct, Huberman says, though he doesn't think that's what Csikszentmihalyi originally intended; the neuroscience says the brain changes only when there's some mismatch between what you're trying to do and what happens [6]. "The brain doesn't have any reason to change unless there's frustration agitation or at least some neurochemical Change associated with those things that we call frustration and agitation," he says [6].

Huberman: the evidence from prism glasses

The key experiments come from Huberman's Stanford colleague Eric Knudsen [1][2]. In the superior colliculus, maps of visual, auditory and motor space are stacked and aligned, so that a sound, sight and movement at the same location line up [1][2]. Subjects wore prism glasses that displaced the visual world, so objects seemed to be somewhere they weren't [1][2]. At first they reached to the wrong place. Young subjects adjusted within a day or two; older subjects took much longer, and some never adjusted [1][2]. Crucially, the signal for the shift wasn't the prisms themselves or the displaced image; it was the errors, the missed reaches [1][2].

Adults could catch up in two ways [1][2]. The first was incremental learning: shifting the world by about 7 degrees, then 14, then 28, so adults faced smaller errors they could tolerate and stack into a large change [1][2]. "Put simply incremental learning as an adult is absolutely essential," Huberman says [1]. The second was contingency: when subjects had to use the displaced cues to find food, adult plasticity became as dramatic as a young subject's [1][2]. "How badly we need or want the plasticity determines how fast that plasticity will arrive." [1]

Huberman: fail many times in the same session

In his episode on learning skills faster, Huberman turns this into a practice rule [3]. Errors cue the nervous system to two things: error correction, and opening the window for plasticity [3]. "Errors tell your nervous system that something needs to change." [3] If you're learning to dance and stepping on your partner's toes, "those errors are opening the possibility for plasticity. If you walk away at that point, you've made the exact wrong choice. Without errors, the brain is not in a position to change itself." [3] Errors recruit frontal cortex networks, top-down processing, and neuromodulators such as dopamine, acetylcholine and epinephrine [3]. He calls it the framing effect: "It frames what's important." [3]

Errors also solve a problem that trips up many learners: where to put attention [3]. "Errors will tell you exactly what to focus on," he says, and "those failures will cue your attention to the appropriate sensory events" [3].

He is careful to say this isn't a pep talk. "This isn't about motivation to learn," he says [3]. "I'm not saying, oh, go make errors. Errors are good for you. You have to fail in order to win. No, you have to fail in order to open up the possibility of plasticity. But you have to fail many times within the same session." [3] His rule: set a block of time and perform the maximum number of repetitions you can do safely within it [3]. In early sessions, "the key is to make many errors, to let the reward process govern the plasticity, let the errors open the plasticity", then let the brain go idle afterward and maximize sleep [3].

His evidence includes the study nicknamed the Super Mario effect, involving about 50,000 subjects guiding a cursor through a maze [3]. When they made a mistake, half saw "that did not work, please try again", while the other half saw "you just lost five points, please continue" [3]. Huberman admits he expected loss aversion to win, based on his knowledge of dopamine circuitry [3]. Instead, the try-again group succeeded 68% of the time versus 52% for the point-loss group, because they made many more attempts per unit time while the point-loss group gave up earlier or entirely [3]. How errors are framed changes whether people keep making them.

A striking detail is his finding on slow-motion practice [3]. He had assumed that learning a kick or punch in ultra-slow motion first would be best, but the data he found say otherwise [3]. Ultra-slow movements don't give accurate proprioceptive feedback, and "it's too accurate. You don't generate errors." [3] They seem to help only once you reach a success rate of about 25 or 30 percent, not at 5 or 10 percent [3]. Practice that feels clean and controlled can deprive a beginner of the very signal they need.

Huberman: strain is the entry point

In his foundational episode on how the brain changes, Huberman explains that deliberate, top-down effort feels like agitation because it involves the release of norepinephrine, and that this discomfort isn't a sign something is wrong [4]. When you're learning something new or trying to be more deliberate, it's supposed to feel challenging; "the agitation and strain is the entry point to neural plasticity." [4] He says "agitation and a feeling of strain are actually required for this process of neural plasticity to get triggered" [4], and that when planning a focused learning bout, "the expectation should be that the early phase of that cycle is going to be challenging" [4]. "It's going to hurt. It's not going to feel natural." [4]

There's a dark side to the same mechanism. "It's actually very easy to get neural plasticity as an adult through traumatic or terrible or challenging experiences," he says, because bad events flood the brain with the alertness and focus chemicals that mark circuits for change [4]. The rewiring itself, he adds, happens later during sleep and deep rest [4].

Huberman: attach dopamine to the errors

Since dopamine responds to what we believe is good for us, Huberman suggests amplifying the effect by telling yourself errors are valuable [1][2]. Combining repeated failure with the belief that those failures are good for learning creates an outsized effect, he says [1][2]. He describes the ideal as scheduling learning at your sharpest time of day, getting to the point of making errors, and then keeping going for 7 to 30 minutes: "just keep making those errors and drill through it and you're almost seeking frustration and if you can find some pleasure in the frustration yes that is a state that exists" [1]. That window, he says, "is exceedingly frustrating but that frustration it liberates the chemical cues that signal that plasticity needs to happen" [1]. The payoff often comes a day or two later, after a nap or a night or two of deep rest, when much of what you got wrong now works [1].

His own example is his free throws, which he describes as terrible at age 45 [2]. He shoots until he hits frustration and then continues for somewhere around 10 to 100 more attempts, letting the motor system teach itself, as long as he doesn't pile on new kinds of errors that confuse it [2]. And the benefit spreads: "Plasticity is a state of the brain and nervous system." [2] After an error-filled session you stay in a heightened learning state for at least an hour, when you could read, study a language or even work on something in therapy [2].

Growth mindset research: how people treat errors decides whether they improve

Huberman's episode on growth mindset explains why some people use errors and others are defeated by them [5]. He describes a brain-recording study by Mangels and colleagues, with Carol Dweck as senior author, in which participants answered trivia questions and were told whether they were right and what the correct answer was [5]. People with a fixed view of intelligence showed a larger error-related signal over a front part of the anterior cingulate cortex, linked to emotional and bodily responses, when told they were wrong [5]. In people with a growth mindset, the error signal leaned toward areas associated with cognitive appraisal [5]. In short, growth-minded people direct their attention to understanding why they made the error, while fixed-minded people mostly feel the sting [5]. Huberman says this is the key to why growth mindset improves performance, something popular side-by-side comparison lists never explain [5]. He contrasts it with advice like "set your ego aside," which offers no process to follow [5]. The process is to deliberately shift attention to analyzing the error, a habit that strengthens with practice [5].

Feedback shapes whether people seek out errors at all. In Claudia Mueller and Dweck's 1998 study, "praise for intelligence can undermine children's motivation and performance." [5] Children praised for being smart later chose easy problems they knew they could solve to keep earning praise, while effort-praised children chose harder problems that could teach them more [5]. On a second problem set, the intelligence-praised group's performance dropped significantly, the effort-praised group's rose significantly, and the intelligence-praised children attempted fewer problems and were more likely to exaggerate their scores [5]. Praising talent teaches children to avoid the very mistakes that would help them grow.

David Yeager's 2022 synergistic mindsets intervention builds the reframe directly into a 30-minute lesson [5]. It tells students that "difficulty struggle and frustration when you're learning something are not signs that you've reached your limits they're signs that you're expanding your limits." [5] It quotes calculus professor Uri Treisman telling students that everyone will struggle and that stress means understanding is deepening [5]. Results included about a 40 percent improvement in self-regard, a 14 percent improvement in passing particularly challenging courses, and more willingness to take on hard challenges long after the intervention [5]. Alia Crum's work adds that people taught that stress can enhance performance improved significantly on hard tasks without any extra practice [5].

Huberman also revises the "mind is like a muscle" analogy [5]. Resistance training gives an immediate pump, a preview of growth, but learning gives no such preview; while you're learning, you feel worse, confused and frustrated, and that strain is what triggers the learning, which is consolidated during recovery and sleep [5]. A better analogy, he says, would be muscles that shrink during training and then rebound bigger [5]. Understanding that helps you read discomfort as learning in progress rather than proof you can't learn [5]. His advice for feedback follows: use verbs, not labels, and when analyzing errors, don't sugarcoat them but focus on the actions that led to them [5].

Josh Waitzkin: the losses that made him

Waitzkin, the former chess prodigy whose childhood inspired Searching for Bobby Fischer, went on to win Tai Chi push hands world championships and became a coach to elite performers in finance, sports and science [6]. He speaks about losing with unusual gratitude. "I lost last rounds of national chess championships and world championships multiple times over and I and those were the greatest moments of my life in retrospect. They taught me The most important lessons of my life. I would never take it back," he says [6]. "The most heartbreaking devastating moments Ultimately were the ones that catalyzed the most growth and they were beautiful." [6]

His stories show how. At about eight, he lost the last round of a national championship by falling into an opening trap; had he won, he thinks he might have associated winning with no pain [6]. The boy who beat him became his best friend [6]. At 17, on first board in the final round of the under-18 world championship, Peter Svidler offered him a draw in a critical position [6]. Waitzkin never accepted draws, but he now believes the urge to release psychological tension crept into his move, and he played slightly over-aggressively and lost [6]. The loss was so painful he didn't study it for two and a half months [6]. When he did, he saw that the right move was to pull his last defensive piece away from his king, because "His attack needed my defense like fire needs fuel to burn" [6]. He didn't grasp the principle then, but years later he won the 2004 Tai Chi world championship final against a bigger, stronger opponent by "harnessing the power of empty space", letting the opponent lean on him and then disappearing [6]. He sees this pattern everywhere: a devastating loss seeds the insight behind a later breakthrough, in sport, business, writing and love [6].

He also studies errors at several levels at once. When he examines a mistake, he looks at its technical, thematic and psychological layers together, and fixing the underlying theme clears dozens of related errors [6]. "I don't believe in compartmentalization. I believe in thematic interconnectedness," he says [6]. And he has trained himself to learn from other people's mistakes. His concept of "fire walking" means learning from others' errors with the same bodily intensity as from your own [6]. A fighter whose arm gets broken never over-extends it again, but someone watching barely learns; through years of biofeedback, intense visualization and personal triggers, Waitzkin primes his physiology so that studying others' worst moments lands with real force, which he believes could multiply a learning curve many times [6]. He adds a chess tip: many blunders are caught instantly after moving, so write your move down before making it, which forces a return to common sense [6].

He believes you have to seek out the conditions that expose weaknesses. "If you're not sparring against people who can expose your weaknesses Then you don't know what they are you don't have the chance to grow," he says [6]. He tries to live at maximum stretch without snapping: "if you're just succeeding all the time then you're not pushing yourself enough" [6]. He praises Marcelo Garcia, perhaps the greatest pound-for-pound grappler, for never holding positions in training but living in the transitions, and for shedding his entire repertoire the day after dominating a championship and winning two years later with something new [6]. He describes the Boston Celtics' approach of "hunting adversity" now that every opponent targets the champions [6].

He also connects this to Dweck. He rejects the prodigy label; from about age six he was always playing people as good or better and had to work relentlessly [6]. People who identify as prodigies grow brittle because they attribute success to innate gifts, and in competition, he says, you can feel where their energy stops and break them [6]. Huberman adds Dweck's finding that such students were more likely to lie about poor performance, including to themselves [6].

Where They Agree, and Where They Don't

The sources agree on the central pattern. Huberman's neuroscience says the brain changes when errors release epinephrine, acetylcholine and then dopamine, and that flow reflects what you already know [1][2][3][4]. Growth mindset research says improvement depends on turning attention toward errors rather than recoiling from them, and that struggle signals expanding limits [5]. Waitzkin's career shows painful losses, studied honestly, becoming the source of later breakthroughs [6]. Huberman makes the link himself in conversation with Waitzkin, explaining that the brain has little reason to change without mismatch, frustration or agitation [6].

They also agree that walking away at the point of frustration is the costly choice [1][2][3]. And they agree that how errors are framed, as "try again" or as lost points, as analysis or as sting, shapes whether people keep making the attempts that produce learning [3][5].

The differences are about dose and care.

How big the errors should be. Huberman's skill episode says to maximize errors early on [3]. But his adult plasticity work says errors should be small and incremental, because adults can't absorb massive shifts all at once, and piling on new kinds of errors confuses the motor system [1][2]. Waitzkin describes living at max stretch without snapping and, from boxing camps, the tension between building a fighter's confidence and exposing every weakness; getting that balance right, he says, is hugely important and not simple [6]. Everyone agrees errors are needed; the skill is finding the level that stretches without breaking.

Whether losses always help. Waitzkin is candid that devastating losses "can Be brutal for young minds and they can destroy people" [6]. He watched nearly all his young rivals quit and fall into crisis, and says he would never put his own children through the pressure he faced [6]. Huberman's framing of errors is more uniformly positive, though he too notes the dark side: trauma wires in easily through the same chemistry [4]. The reconciliation is that errors fuel learning when people can tolerate and study them; overwhelming failure without support can do lasting harm.

Whether outcomes should matter. A popular reading of Waitzkin's The Art of Learning was that process matters and results don't. Waitzkin regrets how that was used: parents told young players results don't matter, and kids saw through it [6]. You have to care enough to be shattered for the process to work, he says; the skill is to be in the rupture while also holding the perspective your future self will have [6]. Huberman's point that high contingency speeds plasticity supports this: caring about the outcome is part of what makes errors potent [1][2]. Growth mindset research, by contrast, cautions against tying identity to results [5]. The combined view is to care intensely about outcomes while treating any single result as information, not a verdict on who you are.

Flow. Huberman says flow isn't a learning state [2], and in conversation with Waitzkin notes that most people equate flow with getting everything right, which he doubts is what Csikszentmihalyi originally meant [6]. He also says there may not be one optimal arousal state; skill comes from learning the transitions between states [6]. So flow isn't bad, it's just where you perform what you've learned, not where most learning happens.

Why It Works

The sources give a clear chain of mechanisms.

Errors trigger the plasticity chemicals. Mistakes release epinephrine for alertness and acetylcholine for focus on the gap between what you did and what you wanted [1][2]. Partial success then releases dopamine, which lets changes happen fast [1][2]. Correct, effortless performance releases none of this, so the brain has no reason to rewire [1][2][6].

Errors direct attention. They point attention at exactly the features that need fixing, solving one of the hardest problems in skill learning [3].

Errors engage top-down processing. Errors recruit frontal cortex networks as well as neuromodulators, framing what's important [3].

The strain is the chemistry. The agitation of deliberate effort reflects norepinephrine release and is the entry point to plasticity [4]. Learning feels worse before it feels better because the strain is the trigger, with consolidation coming during recovery and sleep [4][5].

Mindset decides where attention goes after an error. Brain recordings show growth-minded people shift to cognitive appraisal of errors, while fixed-minded people show a larger emotional response [5]. Only the first path turns the error into learning.

Belief amplifies reward. Because dopamine is subjective, telling yourself errors are good combines the error signal and the reward signal [1][2].

Small errors can accumulate. Adults adapted to large prism shifts when errors came in small steps [1][2].

Accurate practice starves the signal. Ultra-slow practice is too accurate to generate errors, which is why it helps only after a basic success rate [3].

Where It Applies, and Where It Doesn't

The principle applies to motor skills, instruments, dance and sport [3], and Huberman stresses that the error-driven state also supports language, math and emotional learning [1][2]. Waitzkin applies it across chess, martial arts, foiling, investing, business and love [6]. Growth mindset research applies it to school and to any challenging course [5].

There are limits:

• Beginners versus the advanced. Many errors help early; as skill grows and errors fall, attention can shift to specific features of the movement [3]. Slow, accurate practice becomes useful once you succeed about 25 to 30 percent of the time [3].

• Too many or too-large errors. Adults need incremental errors; massive ones don't produce adaptation [1][2]. New kinds of errors piled on top of each other confuse the system [2].

• Emotional harm. Losses can be brutal for young people and can destroy them [6]. Error-driven learning needs a level of challenge people can tolerate, and adults guiding children should watch for the difference between productive struggle and crushing pressure. Anyone whose frustration turns into persistent distress, anxiety or depression should talk to a mental health professional.

• Trauma. The same chemistry wires in terrible experiences easily [4].

• Safety. Huberman's rule is to perform the maximum repetitions you can do safely [3]; errors in physical skills should be ones you can make without injury.

• The right state first. Too much anxiety or too much fatigue keeps you from reaching the starting line; Huberman recommends arriving clear, calm and focused, perhaps slightly alert [1][2].

What To Do With It

1. Relabel frustration. When learning feels hard, remind yourself that the frustration is the chemical signal telling your brain to change, not proof you can't do it [1][4][5].

2. Stay past the urge to quit. When you hit frustration, keep going, for 7 to 30 minutes of sustained errors, or like Huberman's free throws, 10 to 100 more attempts [1][2].

3. Make errors small and focused. Break the skill into small increments and work on one aspect at a time, rather than attempting everything at once [1][2].

4. Frame mistakes as "try again." Use feedback that invites another attempt rather than counting losses; in the Super Mario study it raised success from 52 to 68 percent [3].

5. Analyze each error, not yourself. After a mistake, ask what caused it, and focus on the actions rather than labeling yourself as bad at the skill [5]. Look at the technical, thematic and psychological layers, as Waitzkin does [6].

6. Praise effort, not talent. Teachers and parents should praise specific actions such as persisting and trying new approaches, so children keep choosing harder problems [5].

7. Seek people and settings that expose your weaknesses. Spar, play or work with people who can find your holes, and if you're succeeding all the time, raise the difficulty [6].

8. Study your losses, even painful ones. Waitzkin took two and a half months to face his world championship loss, but the lesson became a later world title [6]. Write down key decisions before acting, to catch blunders [6].

9. Don't retreat to accurate, easy practice too soon. Avoid slow-motion or comfortable repetition until you're succeeding about a quarter of the time [3].

10. Care about outcomes while treating them as information. Let results matter enough to hurt, but hold the view your future self will have of the setback [6]. Then rest and sleep, so the changes the errors started can consolidate [1][4].

Sources

1. How to Learn Faster by Using Failures, Movement & Balance | Huberman Lab Essentials · Huberman Lab
2. Learn Faster Using Failures, Movement & Balance · Huberman Lab
3. How to Learn Skills Faster | Huberman Lab Podcast · Huberman Lab
4. How Your Brain Works & Changes · Huberman Lab
5. How to Enhance Performance & Learning by Applying a Growth Mindset · Huberman Lab
6. The Art of Learning & Living Life · Huberman Lab