This Week in Startups: “Magic School uses AI to help kids learn, not cheat” (E2196)
Date: October 22, 2025
Host: Jason Calacanis (with guest host Alex)
Main Interview Guest: Adeel Khan, Founder & CEO of Magic School
Notable Segments: Equity vesting Q&A with Eric Glyman (Ramp), On-the-Fly Energy pitch with Chris Canetti
Episode Overview
This episode dives into the responsible integration of AI in K-12 education with Magic School, a fast-growing edtech startup. Host Alex (filling in for Jason Calacanis) interviews founder Adeel Khan, exploring how AI can augment—not replace—teachers, streamline administrative work, and drive individualized learning, all while respecting privacy and nurturing student creativity. The episode also features a practical founder Q&A on employee equity vesting and a pitch from On-the-Fly Energy, a company innovating in grid storage with modular flywheels.
Segment 1: Interview with Adeel Khan, CEO of Magic School (00:00–31:56)
1. Magic School’s Mission & Offerings
-
What is Magic School?
Magic School is an AI-powered toolset for educators—“a teacher copilot”—helping teachers design rubrics, lesson plans, and differentiated materials for diverse and multilingual classrooms.- “Our slogan is teachers are magic. AI should not be the one teaching students.” —Adeel Khan [05:40]
-
For Students:
AI-driven experiences guided by teachers, including:- Writing feedback tools that provide instant, rubric-aligned formative feedback.
- ChatGPT-style safe chatbots for support, designed with guardrails to prevent cheating.
- Personalization for special ed and English language learners.
-
For School Districts:
Districts can deploy a customizable, closed “sandbox” instance of Magic School, embedding district knowledge and aligning tools with local policies and standards.- “They get to kind of create their own ecosystem of gen AI tools for their educators.” —Adeel Khan [03:30]
2. Augmenting, Not Replacing, Teachers
-
Teachers retain full control and must actively design content with specific inputs (e.g., standards, instructions), with opportunities for deep customization.
-
“The teacher is still the center of the classroom… AI can be this incredible augmentation to your work.” —Adeel Khan [05:40–06:25]
-
Differentiation: Snap creation of multiple versions of rubrics or materials—by reading level, language, etc.—to meet a practical need given classroom diversity and teacher burnout.
- “Now you can do this in the snap of a finger.” —Adeel Khan [06:46]
3. Adoption & Growth
-
Intuitive UX tailored to teachers’ daily tasks; praised for authenticity and practical utility.
-
Distribution fueled by word-of-mouth; achieved millions of signups without paid marketing.
- “When you build something that people want, they will share it with others.” —Adeel Khan [13:13]
-
Over 6.5M users, with about half in the U.S.
-
Business model: Generous free tier for teachers, upsold to districts who need security, SSO, and integration with local policies.
-
“The primary customer is the district...they want data security and the ability to customize our AI tools.” —Adeel Khan [14:08]
4. Underlying Technology
- Uses OpenAI and Anthropic models; continues to seek improvements—especially on student guardrails, text complexity, and audio tools (with partners like Eleven Labs).
- Customization per region/state; presence of forward-deployed “solutions architects" to embed and adapt tools locally.
- “If we can apply it really deeply and thoughtfully to the district's needs, then we can...change the world.” —Adeel Khan [15:02]
5. AI Literacy and Responsible Use
-
Rapid rise in student AI use—84% of high school students report using it for schoolwork.
-
Magic School advocates for limited cognitive offload for younger students; increasing, but scaffolded, use as students mature.
-
Anecdote of a 10th grader using Magic School as “a thought partner” but self-regulating use to retain creativity and agency:
- “I don't want to give away my thinking too much—my brain’s in this important development stage.” —Student story via Adeel Khan [21:57]
-
Magic School’s 80/20 rule (“let AI handle the initial 80%, you add the final 20%”) is for teachers, not students:
- “That rule would definitely not apply to students.” —Adeel Khan [23:49]
- “They need to be doing 100% of the cognitive lifting.” [23:49]
6. The Future of EdTech AI
-
Next phase: Deeply personalized, contextually aware AI “buddies” for each student, tracking progress over years (with strong privacy/accountability).
-
“How incredible would it be if it just auto differentiated for you, with knowledge of each of your class periods?” —Adeel Khan [28:11]
-
New capabilities expected to launch as soon as January 2026, with ongoing close partnerships with foundational model providers.
-
Strong international organic adoption, e.g., "80% of educators in the UAE" use Magic School without marketing.
-
“I'm not only now less concerned about the state of K-12, I'm actually more bullish about it.” —Alex [31:42]
Notable Quotes & Moments
- Teachers are Magic: “It is also kind of the ethos of the company. We think teachers need to be in control. AI should not be the one teaching students.” —Adeel Khan [05:40]
- On Teacher Burnout: “One of those things that causes burnout is these incredible expectations we hurl upon teachers and don't give them support to…execute on them.” —Adeel Khan [07:50]
- On Word-of-Mouth Growth: “When you build something people want, they will share it with others.” —Adeel Khan [13:13]
- On Student Agency: “What a thoughtful young lady. Like, how incredible it is that you have this much agency to decide how you’re going to use AI.” —Adeel Khan [22:26]
- On the International Impact: “We have reason to believe that like 80% of educators in the UAE are using Magic School.” —Adeel Khan [31:13]
Segment 2: Founder Q&A—Equity Vesting (33:09–38:38)
Key Takeaways
- Vesting Schedule Norms: 4 years is standard; 6 years is rare but not unheard of. More common at established or “brand name” companies who can command longer vesting for more valuable shares.
- When Longer Makes Sense: If you deeply believe the company will appreciate, 6-year vesting can be worthwhile.
- Recommendation for Founders: Keep terms standard to avoid unnecessary negotiation friction; reserve creative perks (e.g., layered grants) for retaining high performers.
- “If you’re not Elon...maybe keep it simple and focus on the product innovation, not legal documents.” —Jason Calacanis [35:13]
Segment 3: Pitch—On-the-Fly Energy, with Chris Canetti (38:38–57:20)
1. Company & Problem
- On-the-Fly Energy addresses grid fragility and the unreliable backup needs for data centers, AI, and critical infrastructure.
2. Solution: Flywheel-Based Backup Storage
- Product: The Rotor Seed—a modular, 10kWh flywheel energy storage device made from carbon fiber composites, scalable from single units to grid scale.
- Benefits:
- Instant, fire-safe backup with long lifespan (20+ years).
- Higher energy return (95% efficiency).
- Avoids issues of chemical batteries (cost, fire, degradation).
- Network effect: units can be pooled as a virtual power plant for grid stabilization.
3. Business Model & Path to Market
- Units offered under monthly subscription, future revenue from grid services.
- Rollout plan: MVP trials in 2026, launch in telecom/data centers in 2027, larger expansion in 2028.
4. Technical Deep Dive
- Carbon fiber allows for higher RPM and more efficient kinetic storage than old metal flywheels.
- Size: 18" diameter/length per module.
- Highly modular; can be stacked in shipping containers or larger installations.
- “I will save a slice [of equity]—I'll put a sticky note on the cap table for you.” —Chris Canetti to Jason Calacanis [56:31]
5. Applications & Market
- Data centers, manufacturing, EV charging, residential backup, and potentially air conditioning grid support.
- Discussion of the “duck curve” and importance of matching intermittent renewal generation (like solar) with energy storage options.
6. Broader Energy Discussion
- Chris emphasizes a mixed, pragmatic path to U.S. energy independence, balancing near-term (natural gas, existing infrastructure) and long-term (renewables + massive storage + nuclear) strategies.
- “America needs energy independence and how do we get there? ... If we can have renewables and be able to take out the downside…that’s the solution I see.” —Chris Canetti [53:11]
Timestamps for Key Segments
- Magic School interview start: [03:05]
- Differentiation & rubrics: [06:25]
- Adoption & growth: [11:46]
- Distribution & business model: [13:12]
- Customization for districts: [15:02]
- AI for students: philosophy: [18:21]
- Student story on AI use: [19:43]
- 80/20 rule clarification: [23:08]
- Magic School roadmap: [25:13]
- International expansion: [30:48]
- Founder Q&A: Vesting schedules: [33:09]
- On-the-Fly Energy pitch: [38:38]
- Flywheel tech, cost, efficiency: [43:00]
- Applications and market fit: [47:05]
- Duck curve, energy challenges: [50:05]
- Vision for America’s grid: [53:11]
- Pitchwind-down and investor interest: [56:22]
Conclusion
This episode articulates the optimistic reality of responsible AI in education, providing a practical, teacher-first toolkit with Magic School, and highlights the potential for scalable innovation in grid storage with flywheel energy. Strategic discussions clarify best equity practices for founders, and real founder pitches keep things grounded in entrepreneurial grit.
Magic School stands as a model for both ethical AI deployment and viral product-led growth in a deeply complex sector.
On-the-Fly Energy showcases how classic ideas (the flywheel!) are being reimagined with modern materials and AI-era urgency.
