Reinventing School for the AI Era | Joe Liemandt, Gagan Biyani & Erik Torenberg

Joe Liemandt describes Alpha as a reinvention of the K–12 school day rather than merely another education technology product. He argues that classroom instruction has changed little in 100 years, even though cognitive science, learning science, and large language models now make it possible to build a “learning engine” that teaches students roughly 10 times faster. Alpha uses that capability to deliver core academics in about two hours per day, during which students learn twice as much as they would through six hours of class plus homework. The remaining hours are devoted to five categories of life skills: leadership and teamwork; entrepreneurship and financial literacy; socialization and relationship-building; storytelling and public speaking; and grit or hard work. Students develop those skills through group projects with their friends and adult coaches. Liemandt says Alpha has grown from 12 campuses to 50, and that test scores, student happiness, customer satisfaction, and workshop quality have improved as the system has scaled. Erik Torenberg says the important achievement is making the model believable: eventually, he expects people to ask why schools ever operated differently. Liemandt agrees that the challenge is now productizing and scaling a high-quality school model, with AI helping Alpha “get better as we get bigger.”

When Torenberg asks why Alpha students report enjoying school more than vacation, Liemandt rejects the idea that the school is simply making learning easy or turning it into entertainment. Alpha surveys found that more than 90% of students say they love school, while 50% say they would rather attend school than go on vacation; at one campus, that figure is 65%. Liemandt’s explanation is that students enjoy doing difficult things with friends and seeing themselves improve. “The key to your child’s happiness is high standards,” he says: children want to do hard things when those challenges feel meaningful and social. He gives teamwork as an example of a skill traditional schools often handle poorly. Ordinary group projects frequently leave one student doing the work while everyone else waits, teaching an “anti-pattern” rather than cooperation. Alpha puts third graders in escape rooms, uses AI to analyze their transcripts, and measures how much of their language is supportive and team-oriented. Students begin with less than 20% uplifting or collaborative language, then complete an eight-week workshop and must reach 85% team behavior to pass. Fourth- and fifth-graders also take the Wharton MBA teamwork simulation, which Liemandt says 50% of MBA students fail. Students are shown that they can finish an escape room three times faster through effective teamwork, turning a demanding lesson into a motivating, visible achievement.

Liemandt says students dislike school partly because they are repeatedly asked to do work at which they are failing. In sports, adults sometimes hold a child back or place them with a different group so they can experience success; traditional, time-based academic systems rarely do that. Once a student falls behind, the class continues moving, and the student experiences the rest of school as “torture.” Alpha’s learning engine instead gives students lessons based on what they actually know. An eighth grader who missed material earlier might receive a fifth-grade lesson, master it, and then move forward without being defined by the gap. The engine aims to keep students in an 80–85% accuracy range: 100% correct means the work is too easy, while falling below roughly two-thirds correct creates disengagement. AI can continuously adjust the difficulty, provide scaffolding when work becomes too hard, and let students move faster through material they have already mastered. Liemandt calls this an “unending stream” of appropriately challenging questions that a teacher facing 30 students cannot provide individually. The result is that students see themselves succeeding academically, while still being held to high standards; their motivation comes not from lowering expectations but from making achievement attainable and visible.

Gagan Biyani says Alpha supports the thesis behind Horowitz and Andreessen Academy: modern tools can redesign education to improve academic outcomes while also producing stronger networks, more agency, better character, and a greater willingness to do difficult things. He cautions, however, that the model cannot simply be copied unchanged across age groups. K–8, high school, and college-age students have different needs, so a student-centered institution must adapt its design at each transition. Liemandt says he became involved with HAA because universities need reinvention just as K–12 schools do, and because he wants Alpha graduates to have somewhere aligned with their preparation. He does not believe university will disappear; he believes it can become substantially better. More than 70 students from MIT, Stanford, and UT spent a summer interning at Alpha, and Liemandt says they were “terrified” that their universities were not preparing them for a labor market in which entry-level jobs may be automated. During the internship, they worked roughly 80 hours per week researching learning science and building an educational LLM intended to be free for people worldwide. They were not assigned textbooks and told to study abstractly; they learned the necessary theory in service of building something real. Biyani argues that ambitious college students should get hands-on experience, make mistakes, and enter the real world before graduation.

Biyani says motivation is central because students are more likely to learn when education is connected to something they want to build, use, or earn from. He recalls finding college’s format ineffective and says even current university campuses appear to have very low attendance in many lectures. Liemandt reports a similar reaction from Alpha’s first graduating class: after entering university, the students were earning 4.0s but described large introductory lecture classes as a “complete waste of time.” Five wrote editorials for their school newspapers criticizing lectures in which students scroll TikTok, use laptops, or do not attend. They still valued small seminars, clubs, activities, and the people they met. Liemandt describes university as a bundle of products whose strongest elements may be social and experiential rather than the formal academic component. He recalls dropping out of Stanford but valuing encounters with people such as Steve Jobs, Bill Joy, and Scott McNealy, as well as building a company with friends. Torenberg and Biyani argue that HAA is designed to make those high-value elements central rather than accidental. Employers, they say, are less concerned about whether graduates know a narrow technical skill than whether they possess maturity, people skills, systems thinking, agency, and the ability to handle rejection and complexity. Those capabilities come from interacting with people and building projects, not from passively sitting in class.

Liemandt says parents are increasingly questioning the conventional route, even when they themselves attended Stanford, Harvard, or another elite university. Kindergarten parents in particular look 12 years ahead and ask whether the current path will prepare their children for a world transformed by AI. Parents of older students often remain committed to getting their children into college because they are already near the end of the traditional process, but younger families are more willing to consider an alternative. He compares the moment to 1910, when roughly 90% of Americans had only an eighth-grade education, more than 30% worked on farms, and the arrival of the Case tractor made many inherited occupations obsolete. Communities that had never experienced high school voted to raise taxes and create schools because they understood that their children needed a different preparation. Liemandt says today’s parents face a similar choice. For families wondering why they should risk a new institution instead of pursuing an Ivy League school, he says adoption follows a curve: Alpha first attracts families who already believe the existing system is failing, and broader demand follows once the first cohort demonstrates that students thrive. He admits that he himself took two years to move his daughters from Catholic school to Alpha because he initially thought a school operating out of a garage was strange. He now says he was slow, and that his daughters still remind him of it. The speakers emphasize that early adopters should be genuinely aligned with the model, because start-ups have inevitable hiccups and highly skeptical families may struggle with them. HAA expects a one-year experience to be less risky because students can defer or reapply to conventional colleges afterward.

The group rejects the question, “Which field should I study so AI cannot replace me?” Biyani says the better question is what a student actually cares about, because an AI-native person can become highly capable in any field. He argues that old-fashioned mathematicians or programmers who insist on chalkboards and pencil-and-paper methods may be displaced, while students who love those disciplines and use AI creatively can accomplish things previously considered impossible. Torenberg similarly says students should approach the future on offense rather than defensively seeking a safe career. Liemandt describes two complementary effects of AI: cognitive unloading, which lets students acquire domain knowledge dramatically faster, and expanded building ability, which gives them “superpowers” to create products and systems. He expects students to write much of the future’s learning software, including tools designed by peers who understand how to make learning engaging. He cites an Alpha student named Austin who built an AP world-history curriculum around a GeoGuessr-style game, historical events, and a “snarky AI”; every student who used it earned a 5 on the AP exam. Alpha’s high-school students, he says, are not pessimistic about AI but tell adults, “You’re cooked. We’re coming for you.” Liemandt also explains Alpha’s Founder School, a specialized high-school program teaching entrepreneurship and financial literacy. Students progress through earning $100, $1,000, $10,000, $100,000, and $1 million. The school guarantees that students will start a business earning $1 million in profit while in high school or receive their tuition back. He expects AI to shrink large companies while producing an explosion of entrepreneurship, and says students should be prepared to build the future rather than wait for it to happen to them.