How to Accelerate Learning & Improve Education | Joe Liemandt
In a Nutshell
Alpha School replaces traditional classroom instruction with AI tutors delivering mastery-based, individualized lessons—students achieve top 1% academic performance in just two hours daily. The remaining day focuses on workshops teaching entrepreneurship, financial literacy, communication, and grit through hands-on projects like starting businesses and organizing festivals. High standards combined with individualized support, plus payment incentives for completing difficult tasks, enable students to break through self-imposed limits and reach performance frontiers regardless of starting ability.
These notes were generated by AI and may contain inaccuracies.
In America, the education system moves students forward without ensuring mastery of foundational skills. If a student gets a B in fourth grade at 80% accuracy, the system says "move on" to fifth grade material rather than addressing the gaps. This is contrasted with sports, where coaches focus on fundamentals like ball handling when players lose the ball 20% of the time, rather than moving to dunks and alley-oops. The problem is that academic gaps compound over time.
Andrew Huberman introduces Joe Liemandt, principal of Alpha School, which ensures every child reaches the frontier of their knowledge and skill potential. Students spend just two hours daily on core academics to ensure no gaps, then spend the rest of the day learning to start businesses, develop financial skills, and build communication and interpersonal abilities. The approach tailors learning to individual needs without letting children pick what's easiest or forcing uniform pacing across subjects.
Alpha School data shows K-2 students test in the top 1% nationally, with average SAT scores of nearly 1530. Many students start successful businesses generating hundreds of thousands to millions of dollars while still in school. The school serves students of all backgrounds and income levels, not just wealthy families.
One hundred years ago, education systems aimed for universal access, importing the Prussian industrialized model that put students in classrooms with teachers at the front. This model remains virtually unchanged across the globe—from wealthy American school districts to poor districts in India. While universal access has been achieved, universal learning has not. There's vast disparity in how much students actually learn during their 12 years of schooling.
New technologies including AI now enable revisiting and fundamentally changing education for the first time in 50-100 years. The goal is to build a scalable model that can reach a billion children.
Parents understand milestones based on their own educational experiences: reading by third grade, memorizing multiplication tables by third grade, learning cursive, reading specific book lists. However, academic standards have declined significantly. Parents of kindergarteners forecast forward 12 years and conclude the current system won't prepare their children for the future, creating demand for educational change.
While online resources like YouTube, Khan Academy, and MIT OpenCourseWare provide access to materials, school serves purposes beyond content delivery. Schools provide socialization, motivation, and structure that self-directed learning cannot replicate. In 10-20 years, 90% of parents will continue dropping children off at buildings with other children and adults.
The tension between coddling children versus pushing them creates a confused educational landscape. Alpha School's philosophy requires students to do hard things regardless of whether they enjoy them initially. The school rejects the notion that children should only do what they love, instead insisting they must master essential skills.
Joe Liemandt initially resisted Alpha School, viewing it as "janky" and "crazy" compared to his Catholic school experience. There was no teacher at the front of the classroom, and students learned through apps while doing handstands. After two years of persuasion from the founders, his daughters attended for one week before summer camp and refused to leave, preferring school to camp. This convinced him to enroll them 10 years ago.
Alpha School makes three commitments to students. The first is that students will love school, with 90% reporting they love school in surveys. When asked whether they would rather attend school or go on vacation, 40-60% chose school depending on the vacation and workshops offered. The second commitment is that students have far more potential than parents expect.
Alpha School maintains that high standards are essential for student happiness. The school is the top academic performer nationally, with every grade level in the top 1% across subjects. Freshman honors students average 1400 on the SAT, exceeding most senior classes. The paradox is that students cannot love school more than vacation without high standards—the tension between high standards and high support creates engagement.
American culture applies different standards to sports versus academics. Parents want their children on championship football teams because it will be the hardest, most memorable experience requiring maximum effort. This sports ethos—if applied to academics—would transform education. Great coaches combine high standards with high support, and this same model should apply to academics.
Drawing from Dr. Joerger's work on motivating children, Alpha School follows the principle of combining high standards with high support. Most parents fall into one of two categories: high standards with low support (throwing children into challenges without guidance) or low standards with high support (avoiding any stress). The effective approach requires both. Child development experts confirm that children develop through struggling and failing on the path to success, supported by caring adults.
Alpha School's workshops demonstrate high standards in action. Kindergarteners climb a 40-foot rock wall, and at the San Francisco campus, they learn to ride bikes and ride 5 miles across the Golden Gate Bridge. During these challenges, children sometimes get stuck halfway up, whimper, or struggle. Parents instinctively want to rescue them, but the guides coach from below. Eventually, children succeed, ring the bell at the top, and come down beaming. Parents witness their children achieving what they thought impossible, creating "fanatical love" for the school.
Traditional societies had rights of passage that were intentionally difficult. The core developmental principle is struggle and failure on the road to success, supported by adults. This happens extensively in sports but rarely in academics. The same loops of challenge, failure, support, and eventual success should operate in educational settings for seven hours daily, just as they do in after-school sports programs.
America holds the belief that anyone can become anything they want, supported by examples of the 10 model. However, this model is not entirely accurate. Joe Liemandt attended a competitive public high school where phenoms existed in sports, including twins Joe and John Cannon. One became a goalie for the San Jose Quakes, while the other had a 100 mile plus fastball in high school and played professionally, possibly for the Mariners. Their younger brother Colt Cannon picked up a skateboard and within eight months was on a team with a pro model. The Cannon family demonstrated excellence in both athletics and academics, though none became professors at Stanford.
In sports, genetics and family lines play recognized roles in determining achievement. Society accepts conversations about these biological factors in athletics. The tension lies between the belief that anyone can become anything and the reality that physical limitations exist.
The concept of support remains unclear. At one extreme, zero support means children figure things out independently, such as making their own breakfast. At the other extreme, excessive support involves doing everything for the child, including climbing rocks for them or completing their homework. Neither extreme produces optimal outcomes. The challenge is defining what truly useful, standardized support looks like.
Athletics demonstrates clear knowledge of how to reach performance frontiers despite physical constraints. Liemandt's daughter aspired to become the world's best volleyball player and start for Stanford, practicing every afternoon. Upon entering high school, she recognized her physical limitation of insufficient height for her position. The key difference between athletics and academics is that sports have established training systems that deliver kids to their performance frontier.
Athletic training does not rely on inquiry-based approaches where students productively struggle without instruction. Coaches provide step-by-step guidance, such as showing exactly how to block or instructing players to execute their position precisely, as Nick Saban directs. Schools, by contrast, often employ inquiry training that lets students figure things out independently, which proves ineffective for teaching sports skills.
Traditional classroom settings present a fundamental problem: teachers deliver median instruction to all students simultaneously. In a football team analogy, this would mean the quarterback and wide receiver receive identical instruction. The teacher must provide the exact same content to 25 or 30 students regardless of individual needs.
The education system provides universal access but not universal learning because students require individualized instruction. Some high schools offer honors tracks that shift the mean by grouping higher-performing students together, but this represents only tiered grouping rather than true individualization.
Individualized tutoring represents the optimal educational model, as demonstrated by Plato, Socrates, Aristotle, and Alexander the Great. The one-on-one mentorship and apprentice models prove most effective. However, these approaches lack cost-effectiveness. A century ago, when education needed to reach everyone, 30-to-1 ratios represented what society could afford. Parents consistently rate schools on class size, with smaller ratios considered superior: 20-to-1 beats 30-to-1, 10-to-1 beats 20-to-1, 5-to-1 beats 10-to-1, and 1-to-1 proves best.
Homeschooling produces bimodal results depending on parental standards. Parents who reject standardized tests and curriculum achieve poor academic outcomes, with students demonstrating low academic knowledge because parents prioritize life skills over academics. Parents who prioritize academics achieve dramatically better results through one-on-one environments, especially when former teachers homeschool their children.
Some homeschoolers attend Alpha schools for afternoons to access high academic standards alongside life skills development. The goal remains achieving one-on-one instruction ratios.
Learning science encompasses cognitive load theory and cognitive science. Bloom's 1980s paper on the two sigma program demonstrated that individual tutors enabling mastery-level learning allow students to achieve two sigma performance, reaching the top 2% of learners. This assumes high motivation levels. Mastery proves essential—tutors without mastery-based approaches fail to achieve top 2% outcomes. One sigma derives from tutoring itself, while another sigma comes from mastery-based instruction. Both elements prove necessary for optimal performance.
AI tutors now enable cost-effective delivery of perfectly tailored lessons at each student's level with mastery-based approaches. This breakthrough allows students to learn 10 times faster than traditional methods. Alpha schools achieve top 1% academic performance in just two hours daily. Students spend an average of 121 minutes in AI applications and learn more than twice what they would through six hours of classroom instruction plus homework, achieving 5 to 10 times normal learning speed.
AI tutors differ fundamentally from chatbots. When chatbots enter standard schools, 90% of students use them to cheat. An Ivy League professor administered an in-person final where only two students failed because others used chatbots to complete their papers without reading assigned books. Chatbot summaries replaced actual reading. AI technology requires proper implementation to avoid negative outcomes.
Learning depends on having appropriate foundations before advancing. Material that exceeds current knowledge levels cannot be learned regardless of effort or mindset interventions. Students receiving MCAT content without prerequisite biology knowledge will fail. The current education system delivers eighth-grade content to eighth graders regardless of their actual skill levels, including students functioning at seventh, sixth, fifth, fourth, or third grade levels due to COVID-related learning loss.
Some students never learned to read in third and fourth grade, yet receive eighth-grade instruction. The removal of memorizing multiplication tables, science of reading approaches, and phonics instruction eliminated foundational skills that become critical in higher grades. These changes contributed to declining performance curves.
Schools of education do not emphasize cognitive load theory, cognitive science, or learning science fundamentals. When Alpha surveyed hundreds of their top teachers, only 5% had learned direct instruction during their education. Project Follow Through, the largest educational study from the 1970s examining effective teaching methods, remained unknown to teachers with master's degrees and PhDs.
A former Houston School District teacher who became an Alpha guide reported learning more about how children learn from a six-hour presentation than during her master's or PhD programs. Education lacks the standardized textbooks common in fields like neuroscience, where Principles of Neurobiology by Kandel, Schwartz, and Jessell or Principles of Neuroscience serve as universal foundational texts across institutions.
Education operates as opinion-based rather than fact-based. A dean from one of the country's best education schools referenced Galileo's trial when discussing the status of facts in education, describing it as a religious war rather than a fact-based trial. The UC system's debate over standardized tests illustrates this pattern: 2,000 professors supported requiring standardized tests, but the board of regents voted against them.
School purposes extend beyond academics. Parents may prioritize specific teachers who changed their older child's life over academic performance standards. When schools face closure for failing to meet academic benchmarks, parents express greater concern about teacher continuity than test scores. Government funding targets academic education, but parental demands often diverge from this focus.
Inquiry learning versus direct instruction represents a major ideological conflict in education. Alpha's approach of teaching students 5 to 10 times faster attracts learning scientists who recognize breakthrough data that previously could not be implemented.
Scientific fields including chemistry, biology, and physics experienced periods of debate and uncertainty, similar to current educational disagreements. Each field advanced through invention of measurement instruments: the light microscope for biology, the telescope for astronomy, and the analytical balance for chemistry. These tools enabled finer-grained measurements that transformed understanding.
AI provides education's equivalent measurement breakthrough by enabling one-on-one instruction with detailed data collection. This creates closed-loop data systems that allow precise measurement of lesson effectiveness through engagement tracking, immediate question responses, and spaced repetition testing for long-term memory transfer. The data loop enables 5 to 10 times learning improvements.
Classroom-based research suffers from contaminated data where effect sizes reflect random classroom variables rather than learning science principles. When testing new approaches in traditional classrooms, implementation varies, prerequisite knowledge differs among students, and motivation levels fluctuate unpredictably. AI environments provide controlled conditions where engagement can be measured precisely and learning outcomes tracked systematically.
Learning science theories debate whether students should struggle or receive direct instruction. Neuroplasticity research indicates that error signals drive the opportunity for change—without errors, no reason exists for neural adaptation. The transcript concludes mid-sentence discussing how brains or bodies require error signals to modify existing patterns.
The error signal is what actually drives the opportunity for change and without errors there's no struggle and therefore without struggle there's no learning. There is no world where you can teach without a kid or an adult feeling some degree of mild agitation. Students might learn to think of that as excitement and curiosity, but there has to be some level of arousal. There must be engagement and attention.
A math problem with no instruction to figure it out is one way to try to learn. The better way to learn, the faster way to learn is to give a worked example where you show them and then progressively remove the parts of the example where they're having to progressively go up and by the end they can solve the problem with no scaffolding. It's a question of scaffolding. Worked examples are a faster way to teach kids than just saying good luck. This represents incremental learning with an example at the front end.
Spaced repetition is based on the concept that everybody has a forgetting curve and they're different by person. The approach is to remind them right before they're going to forget it, and if you do that over a few times then it gets recall from long-term memory so it will be there when you need it in the future.
There has been a debate about whether you need initial mastery that you have to hit it with initial mastery the first time you're learning it, or whether you can not learn it to mastery but because of spaced repetition over time by the third or fourth time you'll get the equivalent of mastery. Mastery is defined as 95% or better proficiency.
This can't be measured in a school because that's too fine grained. Alpha School measured it over the last year and their learning scientists found that you don't need initial mastery. You can do it with spaced repetition. At the end of the year, looking at their scores, in 8th through 12th they crushed it with just worked examples and spaced repetition because once you get into algebra, the multi-line problems, worked examples really matter. But in fourth through 7th, it wasn't as effective because the problems are easier, so you lose the benefit and want more initial mastery. When they saw the scores go down, they added initial mastery back to it, so students get the worked example, the initial mastery, and then spaced repetition.
Initial mastery always matters on everything. Whether it's learning vocabulary words, reading comprehension is a function of vocabulary. The best way to do well on SAT English or to know English is to read lots of books and look up the words you don't understand. With a verbal GRE, the advice is to read a good book and underline every word you don't understand, look it up, and then have a vocabulary. Those who do that end up in the 90th plus percentile.
Good schools who do well on the SAT have a great literature program and a great books program. Alpha School has a specific book list that all the kids read because that's what's going to drive vocabulary which drives those scores and drives knowledge acquisition as well.
In their apps, they can teach vocabulary words, but the problem with flashcarding vocabulary is that you don't know its limits. The difference between happy and ecstatic isn't clear through flashcards. Reading 10,000 pages helps understand the difference between ecstatic and happy and where the boundaries of the vocabulary words are, and that's how you build a rich vocabulary and rich reading comprehension capability.
In their apps now, they can give vocabulary words and show the limits and help educate directly where it's a more effective way than reading the book and going to look it up when you don't know it.
Initial mastery matters on everything. The first dimension is concepts: have you solved it once? Have you done it enough to mastery? The difference is between doing three easy problems or doing 10 hard ones. The 10 hard ones is mastery. The second dimension is whether you can do it a month later. Did you cram the night before? Because cramming is bad. Can you do it one month before, two months before? Spaced repetition covers that.
There's a third dimension that really matters from cognitive science. The way the brain works is you have a certain number of working memory slots and then you have long-term memory. Working memory slots are limited, and you have limited attention. Fluency is a magical concept about brain science. If you memorize something to fluency, when you use it, it doesn't take a working memory slot. Working memory slots are a function of IQ. The best way to be farther on your frontier curve independent of brain chemistry is if something is memorized to fluency.
Multiplication tables illustrate why algebra is hard if math facts aren't memorized. If you know 7 × 8 is 56, that takes nothing. If you have to calculate 7 × 8, you're actually using working memory slots. The problem is there are only a few of them. When in algebra with a multi-step word problem, using slots up for basic calculations means they can't be used for other things. When kids get to algebra and struggle, the issue is not the algebra. The issue is the third grade math facts they didn't learn. The way to fix the problem is not to pound them with more algebra. It's to go fix the third grade problem.
A girl who had a low 700 math SAT kept taking the test and kept getting careless errors. The advice was to focus, but they figured out she didn't know her math facts. They had her go back and memorize her multiplication tables. It took a lot of motivational pushing on their guides to convince someone with a 710 that the problem was a third grade lesson. She couldn't convince her parents either. Her attitude was that she didn't need to do that because she was good in math and was in advanced math. A 710 is not a bad score. She was told it was careless errors, but she was just running out of working memory slots. After memorizing her multiplication tables, she got a 790 SAT.
There's a study showing that if the phone is upside down on a table or in a bag under the chair, performance drops. But if it's out of the room, performance is rescued. It doesn't get better. In every case, the student can focus, but the issue is they don't have enough cognitive space to be able to elaborate on ideas, come up with novel solutions. It's not really a focus issue. It's that the brain is computing consciously and unconsciously.
The number of working memory slots is the number of unique chunks that you can think about. There's a limit, and if you're using them for something, you can't use them for something else. A certain math problem takes a certain number of slots, so you can predict whether someone is going to struggle with a math problem and whether it's too hard based on whether they have fluency or not, even if they know the problem and have been taught the material.
Gifted kids have the most working memory slots, but every gifted kid goes through this in math where they do everything in their head and never write down their work. Everybody tells them they're going to have to write down and show their work, but they resist. Every gifted kid at some point in their education hits their wall because there's some math that you just can't do in your head and you have to start writing down. They haven't learned the good habits of how to do a multi-step problem where they're working it down, and they hit a wall and for the first time they're like they can't do a math problem.
Data on child prodigies shows outcomes are not super impressive. Sometimes within a very narrow domain, but the outcomes are usually not that impressive as adults.
Alpha School is building a school for gifted kids. A lot of the problems in the current school system is they're not challenging and they're boring these child prodigies and teaching them the worst habits in the world. They're giving them 12 years of the worst habits and then even their capability can't overcome that. It's like the super gifted athlete who has every bad habit in the book and then in the MLB is doomed. That's what happens with most gifted kids. They're not challenged, so they sit there and can answer every question through the program versus giving them challenging problems when they're in third grade and fourth grade where they're having to learn all the right skills so that they can progress through.
In the culture, gifted athletes are highlighted a hundred times more. It's better to be the quarterback on the Westlake football team than it is to be the number one academic kid at that school in the US. Alpha School is going to find the 100,000 smartest kids in the country and turn them up to 11. If these kids can learn 10 times faster, they want to get it to everybody.
They started this program with Texas having a voucher program. Rural Texas kids whose schools don't have access are joining this program and crushing through it. The goal is to get these 100,000 kids academically MIT eligible by middle school. That's 1580 SATs, fives on Calc BC, Physics C, AP Lang, those kind of things where these kids have all this potential that's being wasted by being bored, and they should just turn them up. Not every kid can get to MIT eligible by middle school based on the frontier concept, but the ones who can, they should. They can get every kid in America dramatically better than where they are today in academics by getting them to their frontier.
Reading is just fluency. When learning how to read, you're struggling because you don't know the words, and then eventually it becomes fluency and you don't have to think about it and you can crush through it and reading becomes much more fun, much more enjoyable. A lot of this is just not doing the basics. The basics need to be nailed.
If you lose the ball 20% of the time going down the court, the coach is not saying let's work on the dunks and the alley oops or even the layups. They're saying ball handling. Twenty percent isn't acceptable. Fifteen percent isn't acceptable. You can't be 85-80% accurate. In America, getting a B means move on to the next level. If you just finished fourth grade at 80%, you get moved into fifth grade stuff. The answer is no, because then all those problems compound.
If you're struggling in chemistry, an algebra test can tell how you're going to do in chemistry because chemistry is basically algebra with cool word problems. If you don't know algebra, the time to try to learn algebra is not in the middle of a chemistry problem. Algebra is basically fast manipulation of fractions, and fluency on fractions comes from having memorized multiplication and division tables. Math more than anything, but even reading on stuff with vocabulary words, it's all just a layer, and if you don't have the foundation and you just keep pushing kids higher, that's why learning curves in this country slow down.
If you take learning curves relatively, kindergarten by the time you get to middle school and high school, the median student in America doesn't learn anything. The NWEA MAP test is a 300-point scale. The average eighth grader to 12th grader goes up one point from 233 to 234 four years later. They don't learn anything. The reason is because if someone is in advanced math but doesn't know multiplication tables and doesn't know it to mastery, they're not going to be able to progress and it's going to eventually catch up with them.
Before, there wasn't the system, but now an AI tutor can sit there and know exactly what you know through assessment, know exactly what you don't know, give a personalized lesson that says you need this lesson, then hit you to mastery, keep you in that loop to mastery, and then let you go on to the next one.
Last year when starting to open lots of campuses as a high-end private school, kids coming from $50,000 private schools, the best schools in the country, the average student when they transferred into Alpha and took their assessment test were 2.2 grade levels behind. This means two things: don't freak out because the scores are about to be sent showing that an eighth grader can't answer some sixth grade questions, some fifth grade questions even. When it got to writing, it was even lower because everybody's been using ChatGPT to write their papers. Parents are going to freak out thinking their kid is two or three grade levels behind, but they're thinking that takes like 200 hours, 180 school days. The average student can do an entire grade level, one subject between 20 and 30 hours in K through 8. So they can catch up. Do an hour a day of homework to catch up and if you're 3 years behind, that's 90 days.
The number one thing AI could be used for is giving kids individualized lesson plans based on their level to catch them up to grade level, and that would be the single best thing to fix education in America.
If you take an eighth grader, you can't go put them in third grade classes, and if you leave them in an eighth grade class, they can't learn. It's not the teacher's fault because they have to teach eighth grade material and aren't teaching the third and fourth grade stuff and the fifth grade stuff. If these people were given individualized lessons, AI could generate individualized lessons for everybody in America and go catch them up because it's not that hard. It doesn't take that many hours. It takes lots of hours when you're teaching years ahead of their capability. It takes 20 to 30 hours per grade level if they're motivated.
Motivation is 90% of the problem or the solution because you have to be able to motivate the person to engage with these personalized lessons to learn it. The first step is to show them it's tractable. When you're three years behind, you're out. It's too far. A year behind is too much work. When you're 22 hours behind, it's 22 hours a day. It's a completely different message. They're taking kids who are years behind and catching them up because they see it and say they can do an hour of homework a day to fix that problem.
One of the wonderful features is the removal of the shame component. Students aren't thought of as robots. Every kid is behind in something, and most every kid is behind in something. The idea that you're 22 hours behind in something means that gap can be made up not in total isolation but not with peers observing.
A friend who is a brilliant and successful musician described an experience in sixth grade where he got called to the board to do math. Being dyslexic, he wrote his seven backwards and the whole class laughed. An older kid told him that if sixth grade is hard, seventh grade is way harder. He told himself at that moment he was terrible at math. It turns out he's brilliant in music and actually exceptional in math because certain music is very mathematical. Key moments in people's childhood—getting picked last for the kickball team—create psychological components that get embedded in people and can set their trajectory off from finding their frontier in a major way.
Digital tutors that don't shame students are wonderful because students can identify gaps without embarrassment and make up the gap. Motivation is 90% of the solution. When discussing learning twice as much in two hours a day, this is independent of starting position, which is critical for motivation and the no-shaming approach. For an eighth grader who needs third grade material, nobody knows they're in third grade material—they're learning twice as much and showing their rate of lesson learning.
All students care about getting their points. The motivational model is based on wherever they started from. If they're learning 2x, eventually they catch up to the norm. A major problem with the school system is measuring schools based on achievement. There are two dimensions: are kids at grade level, and how fast are they learning. Schools should be measured on the slope of the line—the growth rate—not achievement. This is why everybody talks about selection effects, where good schools are just selection effects because they get the best kids. Schools are supposed to develop kids, not just select them.
Growth rate 2x in two hours is used because it doesn't matter where students are. A school's job is to accelerate their growth on lots of dimensions, whether academic or life skills, as much as possible. Parents should put kids in environments where they'll grow at a rate independent of their starting location. Everybody has strengths and weaknesses—the great academic who's not good on life skills or social skills, or the theater kid. Kids' frontier is great across all dimensions and needs to be untapped.
Guides break through blocks where kids say "I can't do that. I'm not that kind of person. I'm not good at that." The job of a school and educator is to crush that block and show the kid it's not a true statement. AI is not going to replace the role of the human. The best teacher remembered is the one who saw a block and convinced the student they could do so much more. Guides ask "Where's the block on this kid and how do I remove it?"
Three middle school girls considered arts theater girls face an epidemic in America where middle school girls believe they're not good at math and science. This block is cultural—they read on TikTok that they can't do it. There are no learning science papers showing middle school girls can't learn math science. They can learn math science at middle school level at their frontier. The motivation needed to get them there was identified. One daughter was told she wasn't a top 1% math science girl, but a bet was made for a thousand dollars she couldn't get to top 1% because she loves to shop. She filled her shopping cart with the $1,000 and every day did the science problems, eventually getting to top 1%.
Aggressive motivational approaches are used to break through blocks, including paying kids straight up, which is controversial. Kids are paid $1,000 to get to top 1% when they go to Alpha. This shows them they can all get top 1%. Self-determination theory questions whether this kills motivation, but it doesn't. The research shows why. Kids have blocks that they can't do something. Whatever can be done, including paying them, breaks through the block. They then see themselves as a different person, creating intrinsic motivation going forward.
After hitting top 1%, the daughter told her mother it changed how she sees herself. She didn't think she was as smart as her older sister, but now realizes she can do everything her sister can do if she just puts the effort in. The concept is putting things through the loop—try hard things, get them through that motivation, whatever it takes. Have them see themselves as capable because they experienced it. A kindergartener can climb a 40ft rock wall. A middle school girl can say "I can be top 1%." Those three middle school arts theater girls all got 790 math SATs. They didn't have to be paid for that—they realized they could do math.
Showing kids they're capable and breaking through whatever their blocks are—academic or life skill side—is the unlock. Kids will do hard things and grind through it. Sometimes the motivation needs to be decided. Even if they're not intrinsically motivated, it's because they see it as too hard and that's not who they are. The question is how to get through that. It's the same in athletics—getting students to do the drills and get the reps. Money is discussed for two things: getting them to do something hard and break through, and doing daily habits where reps build good habits.
Four years ago as principal, seventh graders were told they could all get 100 on the Texas Star state test. The kids said that's impossible. They needed to hole fill—go backward to third, fourth, fifth, sixth, seventh grade wherever they had holes in knowledge to get to mastery. Parents were approached about hole filling, but American parents have been told their kid always has to be at grade level as an educational philosophy, which learning science doesn't support. Parents didn't want to do it. The motivation problem was that kids and parents said they weren't doing it.
Kids were approached directly and told they could get 100 on seventh grade. They said no. They were offered $100 if they got 100—"100 for 100." They said that's still impossible. The catch was any grade level. They could take a kindergarten test and get $100. The Texas Star starts in third grade. They took the third grade test and got $100. They took the fourth grade test and got another $100. Then fifth grade they got 85. They were asked if they wanted AI to generate lessons they missed to get 100. They calculated how many and said they could do that this week. Then they got 100 on sixth grade and seventh grade.
The approach fulfills the requirement to pay kids $100 per lesson per grade level. Third and fourth grade $200 was "wasted" but problems could be fixed for $500 or $1,000. American education spends $20,000 per kid per year in public schools. That's $240,000 or $4 million educating a kid through their education. Taking all middle schoolers and paying them $1,000 to get to top 1%, or all seventh graders $500 to fill holes, is proposed. By law, public schools can't give out cash to kids.
The question is how that law gets changed and what proxy can be given that kids really want. Instead, they pay on the back end when kids go out into society and hit a wall. The societal optimal thing is just to pay the kid. Parents don't like the pay cash approach and Alpha parents give flack about it, but it's super effective when needed. If everybody wanted to fill holes, this wouldn't be necessary. If everybody understood how important it was, it probably wouldn't be needed. Hole filling is mastering the basics—going back like a good coach would say if you're not good at dribbling, let's go back and work on dribbling skills.
The shift in self-perception is most important. The recognition is "I can do hard things" or "I can do what I thought were hard things." If delivered in the correct context and correct way, learning is possible. Learning in one context has similarities to learning in other contexts—it's going to hurt a little bit but doesn't have to make it hurt a lot. If taught something that doesn't require suffering, that's better than requiring suffering. If it can be taught and it's easy, it should be taught that way as much as possible.
When suffering is good, people want to suffer or suffer when doing things they care about. This goes back to motivation. People who say they have the greatest job in the world have a lot of suffering in that job, but they're happy, fulfilled, and excited. Students very motivated academically suffer and struggle through it because they're doing something. No athlete said they never had to struggle or suffer. They did it in purpose of something—able to do the struggle because they were trying to hit their goals.
In K through 5, kids are fundamentally trying to make their parents happy and follow their parents' lead. As soon as they hit middle and high school, it changes—it's social. Middle and high school is about being accepted by the tribe and socialization. A 10-year-old would never do anything to make their parent upset, but at 15 they'll do something they know their parent hates if it gets them street credit or social acceptance. For parents starting in middle school, there's basically six or seven years in middle and high school to decide one of two things: unlock kids' potential or impose value system on them. Parents try to do both.
A lot of suffering that is ineffective in education is kids saying they're not into this and can be pushed as far as wanted but will push back harder. If instead leaning into things the kid loves or they're focused on, they'll do incredibly hard things. The educational system's design question is how to put those two things together so both can be done—learn the hard things that are important while doing things they love.
A sports academy has kids years and years behind academically, bottom 25% in the country. They'll be the largest recipient of vouchers in Texas for families making less than $65,000. Over a thousand kids aren't excited about academics, haven't done academics for years, middle and high school age. They're learning faster than the best private schools in the country—learning two hours, 2x in 2 hours. The question is how to get them to do that hard thing. First, make learning as easy as possible—don't take that eighth grader age-wise and put them in eighth grade material when their knowledge grade is third, fourth, fifth grade. That's impossible, not hard, not doable. Give them third, fourth or fifth grade lessons through individualized lessons.
Students say "This is now doable. It went from impossible to doable." They still have to spend the two hours, grind through it, learn multiplication tables they never did. The system gives them the ability to do that. For the first time, they're in a lesson that's not too hard or too easy—it's right for them. But they're not into education and don't see what's important about it. What drives motivation is at noon if they do the 2x in 2 hours, if the AI says they've learned enough today because they're on track for 2x of two hours, they get to play basketball.
A high school in Dallas has Jermaine O'Neal, the six-time NBA All-Star, as head of the school. Kids get to go practice—they're the number one basketball team in the country. Those kids say they're going to do the academics because they want to play basketball and be on the best basketball team. They have the best baseball team in the country. Sports are the thousand bucks. This is being rolled out broadly depending on access, how many vouchers are available and granted. Across Texas, kids with vouchers get advanced learning that's the best education possible in America—the fastest, better than best private schools—and at noon go to sports facility playing whatever sport they love. These are gated.
Education used to require a GPA minimum to play on after school sports, which drove a lot, but that doesn't exist anymore and has been watered down. In these schools, no cell phones—they're gated where students have to do academics first. If going to one of these schools, odds of a D1 scholarship double because every college coach says if there are two equivalent athletes, the one who's the great academic, who's ground, who's done hard things, is wanted on the team. That's a differentiator—doubling odds of getting a D1 scholarship.
The approach is waking up every day on hard and struggle and asking how to build a school kids love more than vacation, because that's what's going to drive them to do hard things. Adults have to develop kids and say they can't just play and do all the things they want. There are requirements—behavioral standards, can't act that way. No cell phones sitting next to during academic part of the day. K through 8, no way. As getting into high school, cell phones can be earned because they need to be shown that out in the real world there's going to be no one hiding it, so they should hide it themselves, but they're not forced because they need to be ready.
Adults have to set the standards for kids—this is it, this is what's great performance. Kids want to do it and will go live up to them. There's no trouble on the two hours on the academic side in the sports academy because they're doing what they love and see this all makes sense, seems like a fair trade. Kids will do hard stuff. Parents, guides, adults need to set high standards for kids and they can live up to them if helping them be who they want to be, not just pounding them into whatever square peg is thought they should be.
A high school program was just announced. The three commitments are: love school more than vacation, 2x in two hours, and learn all the life skills wanted. There are five stacks including Leadership and teamwork.
The Alpha school curriculum includes dedicated stacks for leadership and teamwork, entrepreneurship and financial literacy, socialization, relationship building, storytelling, public speaking, grit, and hard work, with progression starting in kindergarten for each skill area.
Kindergarteners climb rock walls as part of their grit development. Second graders run a 5K. By eighth grade, students complete a Tough Mudder, emphasizing teamwork by requiring the entire team to cross the finish line simultaneously rather than competing individually.
A new high school program called Founder School opened this year with the promise that students will earn a million dollars in profit by starting a business, or receive all tuition back. The program is built on the principle that life skills are trainable just like academic skills.
Financial literacy is the number one thing parents request that schools don't teach. All kindergarteners learn to earn, save, spend, invest, and donate. Middle school students run Airbnbs, launch food trucks, and operate vending machines while learning gross margin calculations.
Founder School's academic standards match any school in the country. The program framework is "philosopher builders" rather than creating "crypto bros who are just hacking the world to make money." The book list is analyzed by AI as different from traditional classics lists but equally rigorous.
Students willingly engage with rigorous academics because they get to start businesses and earn a million dollars while doing what they love. Parents are responding enthusiastically, with 150 people attending a recent Founder School event instead of the expected 10-15.
A parent testimonial describes a student who previously skipped school 25% of the time but now wakes his mother up to ensure they're not late for two hours of practice, stating he can't keep the coach waiting.
Youth rebellion around puberty serves an evolutionary purpose. Parents can help by adjusting their frame of excellence when they see their child excel in different dimensions. Students experience "suffering with a purpose" when the purpose is internalized rather than externally imposed.
Joe was not academically driven in high school, doing the minimum work required. He calculated exactly what was needed for an A (90 or above) and did only that amount. He was accepted to Stanford but dropped out after realizing he couldn't motivate himself for classes like Spanish when he was interested in AI and expert systems.
Joe's father was devastated by his academic performance, viewing him as a "total slacker" who wasn't meeting his potential. A Forbes quote captured the father's reaction to Joe dropping out of college.
Eighteen months after dropping out, Joe's father changed his view after seeing his son work 100-hour weeks, sleeping under the desk to make his startup successful. The father realized his son wasn't lazy and that the work ethic was simply being applied to entrepreneurship rather than academics.
High school students complete four-year Alpha X passion projects. One student created a Broadway musical using songs sourced from TikTok, filling thousand-person venues and reaching a million streams on the first day. She learned legal rights, music rights, and navigated conflicts with TikTok singers who had millions of views.
A high school student became the first to serve as lead researcher on a paper published in Nature. The rejection rate for even reviewing a paper is 90+%, with final acceptance around 1% or less. Many scientists never publish in such journals throughout their careers.
The Nature paper involved training an LLM to give teens social-emotional advice, specifically in dating contexts. The study found the trained AI crushed parents at giving social-emotional advice because parents give outdated advice that doesn't account for online environments and modern dating contexts.
One student is marketing Martin Scorsese's latest film. AI students from Open Claw presented at South by Southwest and Claude Con, building businesses that earned $50,000 in 10 days.
Students who would reject impossible-sounding challenges like publishing in Nature in eighth grade can be scaffolded into achieving them through progressive skill-building.
The Alpha philosophy holds that this is the best time in history to be a 5-year-old, with the ability to make the first 12 years of education awesome and unlock potential across all dimensions. Students will love school, learn 10 times faster, and possess more knowledge than their parents.
The philosopher builder framework combines philosophy (understanding humans' role with AI, existential debates, great thinkers) with the builder mindset. Schools must create builders who build physically, create businesses, write books, and make videos rather than just sitting in class.
Middle school students entering Alpha are often caught in consumer dopamine loops from TikTok and Fortnite, causing mental health issues and isolation. The Dream Launcher program breaks these loops by having students create vision boards of who they want to be.
Students identify the intersection of what they love, what they're good at, and what the world values and will pay for. This prevents pursuing passions where they lack skill, since "the world won't pay you for that."
Students track every hour of their week in the 168-hour project, revealing the gap between their vision and daily habits. Many realize their current behaviors won't achieve their goals.
Most of the world operates as consumers. Building something represents the ultimate experience in personal life, family, learning, and business. The gap between what people are told they can do and how they're spending their time needs to be addressed.
The education system was built 100 years ago to create consumers for industrial factories, teaching people to follow rules and sit quietly. This created the current system where scrolling becomes the only fun activity. Alpha uses AI and technology to unlock humanity rather than replace it.
Alpha couldn't call staff "teachers" because they're not teaching academics - AI handles lessons. Guides and coaches develop humans, run students through struggle-fail-supported loops, and help kids thrive. Over 100,000 people applied for these roles, including ex-teachers and world-class athletes.
Using AI in education must be a win for everybody - students, parents, and teachers - rather than a win-loss scenario. AI gives humans time to do what they became teachers for: transforming kids' lives and developing humans rather than grading quizzes.
The Prussian model started education systems to create industrial workers who follow rules. Students were lined up in classrooms to sit and listen, preparing them for factory work. This created the current consumer mindset where scrolling becomes the only engaging activity.
The current educational system creates consumers rather than builders. Students spend 12 years in a system where scrolling becomes the only engaging activity, and they never develop the hands-on experience needed to actually create businesses or projects. Watching YouTube videos or accessing MIT content online doesn't translate to real capability - students need to be taught through workshops, collaboration with peers, and actual building experiences.
Afternoons at Alpha are dedicated to builder activities and life skills development. Students participate in hands-on experiences like rock climbing and starting food trucks. One character-building workshop focuses on the value that "the key to my happiness is contributing to my community." This workshop is implemented across all campuses, with Austin students organizing "Alpha City Limits" - a music festival run entirely by fifth and sixth graders.
Fifth and sixth graders learned instruments, formed bands, raised $25,000 through sponsorships, and organized their own music festival. Students learned advertising concepts and realized they could have charged more for sponsorships. The project demonstrates how students can develop multiple skills simultaneously - musical ability, sales, event coordination, and business development.
Alpha incorporates practical life skills like changing tires into the curriculum. Screen time is minimized to 2 hours daily, with afternoons dedicated to physical, hands-on activities. The approach recognizes that traditional shop classes have largely disappeared and that students need physical engagement beyond screens.
Traditional biology classes involving frog dissections and microscope work provide irreplaceable tactile experiences. Alpha is redesigning science labs to make them more engaging than traditional approaches. Students can see brain structures light up with specific labels, creating experiences fundamentally different from watching content on screens.
Alpha is working with Mark Rober to transform science education. Students complete 11 minutes of app-based learning to reach mastery, then spend 45 minutes on engaging experiments and projects. This combination ensures both academic mastery and genuine engagement.
Alpha teaches lab skills through progressive projects. Students begin with cooking and baking to learn following instructions and understanding margin of error. Sixth graders participate in "PFX" (special effects chemistry), involving a 40-step process to create safe breakaway glass. Students become highly engaged, often staying until 6 PM to complete projects for social media content.
Students learn programming and AI through building projects they care about, such as music videos and movie trailers. A weekend hackathon teaching Cursor usage saw students voluntarily working from Friday evening through the weekend. One group of girls built a "Love Island" social app, demonstrating that when projects align with student interests, they willingly tackle difficult technical challenges.
Students developed an app teaching teens that "the key to their happiness is contributing to their community," incorporating daily contribution habits. The project attracted influencer participation, providing both reach for influencers and motivation for students to work on weekends rather than attend parties.
With AI enabling 10x faster learning, the critical question becomes what skills society wants students to develop. Alpha prioritizes entrepreneurship, storytelling, and financial literacy as essential life skills. The goal is shifting from an education system that produces consumers to one that develops builders and creators.
The product name "Time Back" reflects giving students their time back. Traditional education consumes 6 hours daily plus homework. Alpha's system reduces academics to 2 hours, returning time for builder activities and life skills development. This time recovery serves as the primary motivator for students.
A fourth grader at the New York campus who loves quantum mechanics started "Quantum Plations" podcast to connect with other students sharing this interest. This exemplifies how students thrive when given opportunities to pursue their passions and develop as creators.
Alpha appears to eliminate traditional gender performance gaps. ADHD boys who struggle in conventional settings perform well when afternoons involve building activities they enjoy. The system accommodates different learning styles through engaging, motivating content rather than traditional sedentary approaches.
Parents are asked whether ADHD-diagnosed children can focus on video games for 25 minutes. When they confirm this ability, Alpha implements 25-minute pomodoro blocks with engaging content. The approach recognizes that ADHD students can focus when material is sufficiently engaging and motivating.
One student identified as "Rat" completed academic work quickly then built what became Texas's best bike park as a mountain biking project. The park received coverage in Texas Monthly and attracted attention from the Walton family for a similar project in Bentonville. SpaceX subsequently hired him to build a bike park at Starbase, demonstrating how Alpha's approach enables students to develop significant real-world projects.
Alpha started as an expensive private school, following the typical product development pattern of beginning with a viable market before expanding accessibility. The approach mirrors Tesla's progression from Roadster to Model 3, recognizing that educational innovation requires starting with paying customers before achieving broader reach.
Alpha benchmarks against the best educational systems rather than lower-performing alternatives. Their 1400 freshman year achieves 2x learning in 2 hours compared to top 1 percentile students in traditional systems. This competitive approach isolates the effectiveness of their methodology beyond selection effects.
Family income remains the strongest predictor of educational performance in the current system, superseding IQ as the primary factor. Within income levels, IQ and Big Five conscientiousness determine success in traditional time-based rather than mastery-based systems.
Alpha solves IQ-related learning barriers by providing material at appropriate difficulty levels. Students struggling with seventh-grade content can succeed when given fourth-grade material. The US Common Core curriculum is designed so conceptual leaps aren't prohibitively difficult when material is properly leveled.
For students lacking natural conscientiousness, Alpha implements external motivation systems including payment for completing lessons. The goal is reaching a billion students without relying on income, IQ, or natural conscientiousness as success factors.
Alpha is expanding into public schools with average per-student spending of $20,000. Current implementations include schools at $195 and $24 per student. Workshops are being redesigned for these budget constraints, with sports programs being particularly cost-effective due to inexpensive field access and coach-to-student ratios of 1:30.
Simply adding edtech to existing school structures won't achieve desired outcomes. The entire school day must be redesigned to incorporate both accelerated academics and afternoon builder workshops. Some public schools are implementing 3-hour app sessions rather than 2 hours while maintaining Alpha-designed afternoon workshops adapted for their specific constituencies and price points.
Alpha School designs different school types for different student populations rather than trying to create a one-size-fits-all model. The sports academy school targets students who love sports, with the principle that family income doesn't determine access when the product is designed for appropriate price points. Lessons are provided at the right level for motivated students regardless of income background.
The gifted talented school requires cognitive testing for the top 10% of students, targeting the 5% of the market where students want more academics. The approach is to design specifically for these students and provide intensive academics because they love it. The speaker's oldest daughter exemplifies this type of student, reading calculus books for fun and pursuing an English-math double major at Stanford with a desire for more academic challenges.
Schools function as communities where socialization significantly impacts motivation. Placing gifted talented students together creates positive peer effects where students build on each other's capabilities. The number one motivator for overcoming learning blocks isn't teacher instruction but watching classmates succeed at tasks. This peer modeling effect is powerful in athletics where being around higher-performing peers pushes individuals to work harder.
The speaker references their experience as a postdoc at Stanford where being surrounded by peers submitting to top-tier journals created an environment where everyone pushed harder and thought more carefully about research decisions.
There's a cultural tendency to celebrate underdog success stories while showing resistance to people who had advantages achieving their potential. This creates a tension where success without visible suffering or disadvantage is sometimes viewed negatively. Alpha School takes the opposite stance, focusing on developing everyone to their individual frontier regardless of starting circumstances.
The discussion addresses how to ensure access isn't limited by financial barriers or mindset barriers, recognizing that specific circumstances of outlier success stories like Spud Webb aren't easily exportable to broader populations.
Alpha School has committed significant capital, including a billion dollars, to ensure the model isn't limited to high-end private school students. The sports academy model targets families under $65,000 income, with plans to become the largest school serving this demographic in Texas. Texas vouchers enable families with learning disabilities or incomes under $65,000 to access these programs.
Within 100 days of launching a website for Texas Sports Academy when vouchers opened, over 120,000 families expressed interest. The school received slightly over 1,000 vouchers, making it the largest voucher recipient in the state. Students from lower-income backgrounds will use the same academic system as Alpha's high-end private school students, learning at twice the standard rate in two hours daily.
Alpha School is developing partnerships with public schools where students will access the Alpha system during two-hour academic blocks, with afternoons dedicated to workshop-style learning. While public school budgets are lower ($20,000 versus Alpha's $40,000+), the workshops will still provide valuable life skills. MIT's Blueprint Labs will conduct randomized controlled trials to validate the approach and identify what works and what doesn't.
The goal is reaching all 55 million American students, with data collection planned over the coming year to support expansion. The GT squared program aims to make 100,000 students MIT-eligible by middle school, targeting rural students and making the program available nationwide.
Alpha School is opening locations in Africa with a target price point under $1,000 per student. Alpha high school students are building a school in Africa, addressing challenges including solar power, Starlink connectivity, and well drilling while adapting to different community contexts. Students signed agreements acknowledging this would be the hardest year of their lives, motivated by contributing to their community.
A team of world-class video game developers is creating a game that embeds learning science concepts from Alpha's timeback system. The economic model targets 500 million students worldwide, starting with third through fifth grade reading and math before expanding to the full curriculum through 12th grade. The game will be free and designed to replace regular video games with academic content that puts users in the top 1% academically.
Alpha students have tested the game and described it as highly engaging. The motivation problem is addressed by working with influencers and game developers to create content students genuinely want to engage with.
The current period represents a renaissance in education, described as the best time in history to be a five-year-old. Parents will be able to see their children thrive at levels previously unexpected. The approach requires substantial capital investment, with hundreds of millions raised to build these systems.
The speaker describes their four years as principal as the best of their life, emphasizing that changing education is the most rewarding career choice available. The invitation extends to people from all backgrounds, including software engineers and game designers who have never worked in education before. The applicant pool is described as having grown significantly following the discussion.
Keep Andrew Huberman in your library
Save the videos and channels worth coming back to, and find them again in one place.
