LEADERSHIP VOICES

How AI + Incentives Can 10X Learning & Save American Education
with Tech Legend Joe Liemandt

Alpha School principal Joe Liemandt describes a K‑12 model that uses an AI tutor to complete academics in ~2 hours/day, with students learning more than twice as much as peers in six hours plus homework; he claims Alpha is the best‑performing academic school in the U.S., with classes at the top 1% and the ability to move students from the bottom 25% to the top 25% in two years. The core unlock is motivation (“90% of the solution”), delivered through Time Back (finishing faster to do projects students love) and incentives (e.g., paying students, gift‑card unlocks, screen‑time trades). Liemandt says he has committed $1B to scale the model, aims for on‑device AI on a sub‑$1,000 tablet to reach a billion kids, and argues against policies that would ban AI in schools. (blog.joelonsdale.com)

  • Two‑hour academics via AI tutoring + learning science. In Alpha’s model, students spend two hours with an AI tutor instead of a teacher lecturing in front of a class; Liemandt says in that window students “learn over twice as much as kids who sit in class for six and do homework,” and he claims Alpha’s classes are all top‑1%, calling it “the best performing academic school in the country.” He also says the system can move bottom‑quartile students to the top quartile in two years.
  • Five pillars for a reinvention of school. (1) Kids must love school—at Alpha, 96% say they do, and 40–60% say they’d rather attend school than go on vacation; Alpha High opened year‑round at students’ request. (2) Learn 10× faster with AI + learning science versus the 200‑year‑old lecture model. (3) Use the rest of the day for life skills. (4) Adults as Guides (mentors/motivators), not academic lecturers. (5) Community/character/culture as an explicit design goal.
  • Motivation is 90% of the solution; “Time Back” is the #1 driver. Liemandt repeatedly emphasizes that motivation dwarfs ed‑tech itself. Alpha’s product is called “Time Back”—finishing academics quickly so students can do work they love. Guides spend their day on motivation and emotional support, not teaching seventh‑grade science.
  • Incentives: pay the kids; design daily habits. Building on Roland Fryer’s experiments (Houston ISD) on paying teachers, parents, and students, Liemandt says “paying the kids is by far the most effective.” Alpha also runs public‑school pilots for bottom‑10% (MTSS level 3) students where gift cards unlock upon completing lessons, which teachers and parents reported transformed students’ lives.
  • Global learn‑and‑earn pilots. Alpha middle‑schoolers launched a “learn‑and‑earn” program with Ukrainian refugees: $2.50/day for completed lessons, doubled for five‑day streaks; 1,000+ refugees have participated.
  • Cheatbots vs Chatbots; design for integrity. Liemandt warns that deploying general chatbots to students yields “90%” using them to cheat; Alpha instead uses AI as a tutor/coach and argues AI should not be banned in schools.
  • Scale plan & policy. Liemandt states he has taken $1B from his companies to fund a full‑stack reinvention and aims for on‑device AI on sub‑$1,000 tablets to reach a billion kids. He and Lonsdale also discuss school choice and state AI policies (with Lonsdale asserting New York State currently bans any use of AI in schools). (blog.joelonsdale.com)

Key Takeaways & Highlights 🎯

  • Two hours > six hours + homework (on learning gains). In Alpha’s two‑hour AI‑tutor model, students “learn over twice as much” as peers in the traditional six‑hour day plus homework.
  • Top‑1% academic performance & catch‑up claims. Liemandt claims Alpha is best in the country academically, with all classes at the top 1%, and can move students from bottom 25% to top 25% in two years.
  • Kids must love school (measured). 96% of Alpha students say they love school; 40–60% prefer school to vacation; Alpha High runs year‑round at student request.
  • Motivation is 90% of the solution. Guides focus on motivation and emotional support; Time Back (finishing early to do meaningful projects) is the most powerful motivator.
  • Paying students works best (Fryer → Alpha). From the Houston experiments to Alpha’s own practice, paying kids—structured for daily habits and outcomes—produces the strongest effects.
  • Public‑school pilots for bottom‑10%. With MTSS level 3 students, gift‑card unlocks on lesson completion led parents/teachers to report life‑changing results.
  • Learn‑and‑earn with refugees. $2.50/day incentives (double for streaks) helped 1,000+ Ukrainian refugees learn as fast as Alpha’s $40,000 private‑school students, per Liemandt.
  • Cheatbot, not chatbot.” Liemandt says 90% of students use open chatbots to cheat; Alpha’s AI is designed as a tutor/coach, not an answer‑spitter.
  • $1B commitment; on‑device AI < $1,000; billion‑kid ambition. Liemandt describes a $1B funding commitment and a goal to deliver on‑device AI on sub‑$1,000 tablets to a billion learners.
  • Policy context. Lonsdale asserts New York State currently bans AI in schools; both argue for allowing AI and supporting school choice to enable innovation.

Key People & Concepts

  • Joe Liemandt — Principal of Alpha School; founder of Trilogy/ESW; leading the education push described here.
  • Joe Lonsdale — Host of American Optimist; raises policy themes (e.g., AI bans; school choice). (blog.joelonsdale.com)
  • Roland Fryer — Economist whose Houston incentive experiments inform Alpha’s view that paying students works best when designed for daily habits.
  • Time Back — Alpha’s product philosophy: finish academics quickly to earn back time for meaningful projects; the primary motivator.
  • Guides (not lecturers) — Adults focus on motivation/emotional support during/around the academic block and life‑skills work.
  • MTSS Level 3 — Bottom ~10% of students academically; Alpha pilots tied gift‑card unlocks to finished lessons.
  • “Cheatbot” vs. Tutor/Coach — Liemandt’s framing: open chatbots drive cheating; Alpha’s AI is built to teach and coach.
  • On‑device AI (<$1,000) — Scale plan to deliver AI tutoring affordably worldwide.

Share

Listen

Watch

Chapters
  1. Origins: Trilogy → ESW → Alpha. Liemandt recounts building Trilogy/ESW and shifting to K‑12 as his “phase two.”
  2. Fryer’s incentives research & the Houston trials. Paying teachers, parents, and students was tested; paying students worked best when structured for daily habits and outcomes.
  3. The 5‑pillar reinvention (love school; 10× learning; life skills; Guides; culture). Includes measured “love school” outcomes and two‑hour AI‑tutor model.
  4. The learning engine & claims on results. Two hours of AI‑tutor time yields >2× learning vs. six hours + homework; top‑1% classes; bottom‑to‑top quartile in two years.
  5. Motivation architecture (“Time Back,” screen‑time trades). Product is Time Back; examples include screen‑time trades (e.g., 1 hour tutor → 1 hour games) approved by parents.
  6. Public‑school pilots (MTSS‑3) & gift‑card unlocks. Bottom‑10% students with gift‑card incentives completed lessons; parents/teachers reported transformations.
  7. Learn‑and‑earn abroad. Middle‑schoolers organized a program for Ukrainian refugees with $2.50/day (double with streaks), 1,000+ participants.
  8. Integrity & AI design. Avoid general chatbots (“cheatbots”); design AI as a tutor/coach; argue against bans on AI in school.
  9. Scale & capital. $1B funding; plan for on‑device AI on sub‑$1,000 tablets; goal to reach a billion kids over the next 20 years.
  •  

Listen to the Full Podcast

FAQs

Students work one‑on‑one with an AI tutor (no teacher lecturing) and, according to Liemandt, learn over twice as much as peers in six hours plus homework; he claims Alpha’s classes are top‑1%, and catch‑up from bottom to top quartile can happen in ~two years.

The product is called Time Back (finish earlier to do projects you love). Other tools include screen‑time trades (e.g., 1 hour tutor → 1 hour games) with parental buy‑in, and financial incentives where appropriate.

Liemandt cites Roland Fryer’s work (e.g., Houston) and says paying kids—structured to build daily habits—was most effective among teacher/parent/student options.

In MTSS level‑3 pilots (bottom ~10%), Alpha tied gift‑card unlocks to finishing lessons; teachers and parents reported the approach transformed students’ lives.

Alpha’s learn‑and‑earn program for Ukrainian refugees used $2.50/day incentives (doubling with streaks), with 1,000+ children participating.

Liemandt warns that open chatbots become “cheatbots” (he says 90% of students will cheat if given them). Alpha’s design avoids that paradigm and uses AI to tutor/coach rather than provide answers.

Liemandt says he has committed $1B to a full‑stack reinvention and aims for on‑device AI on sub‑$1,000 tablets to reach a billion learners over the next 20 years.

Lonsdale asserts New York State currently bans AI in schools and frames a national debate about whether AI will be allowed to help kids; both discuss school choice as a path for innovation. (This is reported as their on‑air statements.)

Students are taught to steelman both sides of debates; Liemandt recounts formative experiences arguing positions he disagreed with to build understanding.