The AI Math Guy Podcast

About The AI Math Guy Podcast

Where superintendents reimagine education in the age of AI. theaimathguy.substack.com (https://theaimathguy.substack.com?utm_medium=podcast)

Publisher

Jae Kim

Category

education · technology

Language

en

Explicit

No

First episode

23 Jan 2026

Latest episode

21 Feb 2026

Latest episodes

14 episodes in the feed.

1. 

   21 Feb 2026

   Ep. 14 The Spirit Work (w/ Nationally Recognized Superintendent, Michael Matsuda)

   Welcome to another episode of the AI Math Guy podcast! Today, host Jae Kim sits down with educational visionary and personal mentor, Michael Matsuda. After nearly 12 transformative years at the helm of the Anaheim Union High School District (AUHSD), Michael is officially retired—but his work shaping the future of education is far from over. In this insightful and reflective conversation, Michael shares his hard-earned wisdom on school redesign, empowering teachers, and preparing students for a rapidly changing, AI-driven world. Whether you are an educator, administrator, or just passionate about the future of our youth, this episode is packed with paradigm-shifting ideas. In this episode, we cover: * Life After Leadership: Michael opens up about his retirement, the reality of the “superintendent lifestyle,” and his vital advice to young educators about work-life balance and not marrying your job. * “The Future of Public Education”: Get an inside look at his upcoming book co-authored with educational heavyweights Michael Fullan and Barnett Berry. Learn how AUHSD became a national lighthouse for innovation. * Systemness & Spirit Work: Discover the concepts of “connected autonomy” and why building relational capital (spirit work) is the secret glue to making systemic, innovative changes in large organizations. * Predistribution vs. Redistribution: Michael discusses his latest work with Harvard using big data, the Community School construct, and why federal policy must shift to actually break the factory model of education and build real social capital for students. * Navigating the Age of AI and AGI: As Artificial General Intelligence looms, how do we prepare students for an unpredictable job market? Michael explains why the ultimate safeguard against a dystopian future is doubling down on what makes us uniquely human: empathy, relational skills, and love. (Note: We experienced a minor technical glitch at the very end of the recording that cut off Michael’s final thoughts on civic engagement, but the rest of the conversation is fully intact and full of value!) This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit theaimathguy.substack.com (https://theaimathguy.substack.com?utm_medium=podcast&utm_campaign=CTA_1)

2. 

   21 Feb 2026

   Ep. 13 The Rise of the AI Teacher-Builder (w/ Megan Hsu, 2x Hackathon Winner)

   Are you waiting for tech companies to fix the broken systems in education, or are you ready to build the solutions yourself? In this incredible episode, Jae sits down with Megan Hsu—a Stanford-educated Special Education Program Manager with 15 years of experience—who turned her daily frustrations with outdated school tech into a passion for app development. After feeling the weight of burnout and “doomscrolling,” Megan discovered a renewed sense of purpose in AI. Now, she’s a two-time hackathon winner building life-changing tools for neurodivergent students. Whether you are a teacher feeling overwhelmed by tech policies, an aspiring builder, or just someone looking for a spark of hope in the AI space, this conversation will inspire you to stop talking about the problems and start coding the solutions. What You’ll Learn in This Episode: * The Rise of the “Educator Builder”: Why the best ed-tech solutions must come from the insiders who intimately understand the daily hurdles of the classroom. * Vibe Coding vs. Coding with AI: Jae and Megan unpack what it actually takes to prototype and build a functional app, and why “vibe coding” is evolving into true AI partnership. * Winning Apps with Real Impact: Discover how Megan built an AI Social Stories app to help autistic students navigate daily transitions, and Beast Blend, a creative app born out of a single student’s love for mashing up animals. * Moving Past the “Privacy Panic”: Why obsessing solely over AI’s dangers is holding schools back, and how we can pivot to modeling responsible AI use for students. * The Educator Hackathon Vision: Megan’s pitch for bringing teachers together to learn, build, and solve local educational problems in real-time. Memorable Quote: “The best time to start was yesterday, but the second best time is today... You have to fail to learn. You have to fail to improve.” – Megan Hsu This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit theaimathguy.substack.com (https://theaimathguy.substack.com?utm_medium=podcast&utm_campaign=CTA_1)

3. 

   21 Feb 2026

   Ep. 12 AI is the Anthesis of the "Factory Model." (w/ Dr. Beth Fisher)

   Are our schools still operating like 19th-century factories? In this episode of the AI Math Guy podcast, host Jae sits down with Dr. Beth Fisher, an educator and administrator with 18 years of experience, to unpack the systemic challenges of the American public school system and how Artificial Intelligence might be the ultimate disruptor. Dr. Fisher shares her unique perspective from the trenches of an independent study charter school, where she oversees academic accountability and intervention. We dive deep into why students need to unlearn the “game of school,” the transition from using AI as a quick productivity hack to a genuine critical thinking partner, and how to make math matter in the real world through multidisciplinary, portfolio-based learning. The episode concludes with a powerful and emotional tribute to a late colleague, reminding us all of the profound, human “why” behind teaching. In This Episode, We Cover: * The History of American Public Schools: How the industrialization era shaped our modern “One Best System” and why age-grade placement restricts true mastery. * AI as the Antithesis of the Factory Model: Why AI is forcing us to shift from prioritizing productivity to prioritizing human creation. * The Gradual Release of AI Responsibility: Why we can’t just assume tech-savvy kids know how to use AI ethically—and the step-by-step process of teaching them to prompt, reflect, and iterate. * Multidisciplinary Math: Moving beyond memorizing rules and plotting abstract points. Discover how Dr. Fisher’s students use math, history, and science to answer big questions like, “Is growth always good?” * The Future of Education Policy: The legislative shifts required to truly integrate AI and empower students. * Making the Impossible, Possible: A touching reflection on grief, the human-in-the-loop, and the life-changing impact of dedicated teachers. About Our Guest Dr. Beth Fisher is a Director of Academic Accountability at a California charter school with 18 years of experience spanning across teaching, administration, and school foundations. She is deeply passionate about equity, access, and utilizing innovative tools like AI to meet students where they are and close learning gaps. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit theaimathguy.substack.com (https://theaimathguy.substack.com?utm_medium=podcast&utm_campaign=CTA_1)

4. 

   21 Feb 2026

   Ep. 11 Is College Worthless? (W/ Leeland Zhang, Cofounder)

   Welcome to another episode of The AI Math Guys! In this episode, host Jae Kim sits down with Leeland Zhang, the innovative co-founder of Grassroots Academy. Leeland shares his bold journey from studying at UCLA and Northwestern to taking the ultimate leap of faith: dropping out to build a personalized AI learning platform. Together, Jae and Leeland dive deep into the real-world impact of Artificial Intelligence on our educational systems, the evolving job market, and what the classroom of tomorrow will look like. Whether you are an educator, a student, or a founder, this conversation challenges the traditional schooling mindset and explores how technology can bring out the best in human potential. What You Will Learn * The College Dilemma: Why AI advancements and the shifting job market are causing students to rethink the traditional four-year degree. * AI as the Instructor: How Artificial Intelligence is uniquely equipped to handle material-based teaching and grading. * The Evolving Role of Teachers: With AI handling the heavy lifting of instruction, discover how educators can pivot toward relationship-building, mentorship, and project-based learning. * Cultivating Agency: Leeland’s personal advice on finding your purpose, taking ownership of your learning, and building meaningful, real-world projects. * Reimagining the System: A thought-provoking discussion on moving away from standardized testing and toward a more equitable, passion-driven educational model. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit theaimathguy.substack.com (https://theaimathguy.substack.com?utm_medium=podcast&utm_campaign=CTA_1)

5. 

   23 Jan 2026

   Ep. 10 "We don't have much time" (w/ Author, Speaker, & Agentic Expert, Kunal Dalal)

   In our latest conversation, we sat down with Kunal, a former principal and author who is thinking miles ahead of the current EdTech curve. While most of the world is still figuring out chatbots, Kunal is already architecting the next phase: Agentic AI. He describes our current situation as an “incredible fragile moment.” Just like the “golden hour” after a car accident where the body hasn’t yet molded around the injury, we have a short window to shape AI before it hardens into a system we can no longer change. Here is what you need to know about the shift from Chatbots to Agents, and why being “Present Ready” is more important than being “Future Ready.” The Shift: From Chatbots to Agentic Systems We often use the terms “Agent” and “Agentic” interchangeably, but there is a distinction. * An Agent is a piece of software (powered by an LLM) that performs a specific task. * An Agentic System is an ecosystem where multiple agents work together to complete complex workflows with minimal human intervention. The Paradigm Shift: We are moving from Human-in-the-Loop to Human-building-the-Loop. In a chatbot world, you are the trigger. You prompt the AI, check the output, and refine it. You are the bottleneck. In an Agentic world, you become the Supervisor. You build a team of agents—Kunal suggests a “swarm” of about five—to check each other’s work, debate the best output, and execute tasks autonomously. You step out of the machinery and only step back in to audit the results. The “Swarm” in Schools Imagine a school district that owns its own data rather than locking it away in third-party vendor silos. Kunal described a scenario where district leaders currently spend two weeks analyzing grant data to see if it met standards. With an Agentic Swarm, you could point five agents at your local data server and ask: “Analyze the attendance data for 9th-12th grade and tell me what questions I should be asking.” The swarm could return that analysis in 10 minutes. This isn’t just about efficiency; it is about Data Sovereignty. By moving away from massive, energy-hungry cloud models toward local, purpose-built agent swarms, schools can regain control over their information and reduce their environmental footprint. Parenting the AI Kunal’s most profound insight wasn’t technical; it was philosophical. He warns that big tech companies do not have global human welfare as their primary directive—their goal is profit. If we passively accept the tools they hand us, we repeat the mistakes of the social media era. We must “Parent” AI. Just as we guide a child, we must actively shape how AI develops in our institutions. We cannot afford to be passive consumers. “Obsessing over ‘Future Readiness’ is actually abdicating your responsibility... That means someone else has decided your future for you.” Instead, Kunal calls for leaders to be “Present Ready.” This means engaging with the technology right now, understanding its impact today, and being loud about the boundaries and hopes we have for it. The Return of Joy Why go through the trouble of building these systems? It isn’t just to process emails faster. The ultimate goal of Agentic AI is to automate the digital drudgery that keeps us staring at screens so we can return to the one thing AI cannot do: Human Connection. If we can automate the paperwork, the grant reporting, and the data analysis, we can return to the joy of simply being with young people. That is the promise of this technology—if we are brave enough to shape it. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit theaimathguy.substack.com (https://theaimathguy.substack.com?utm_medium=podcast&utm_campaign=CTA_1)

6. 

   23 Jan 2026

   Ep. 9 Neuroscience & AI (w/ Associate Director of AI Labs @BiolaUniversity, Joseph Hartano)

   We recently sat down with Joseph Hartano, the Associate Director of Biola University’s AI Lab. Joseph helps students and faculty navigate the complex intersection of technology, ethics, and theology. Our conversation went deep—from neuroscience to the surprising reason why your AI chatbot might be a sycophant. Here are the key takeaways you need to know about how AI thinks (and why it might be lying to you). 1. The Three Pillars of AI Literacy Joseph introduced us to Biola’s framework for AI education, which is built on a triangle of three essential pillars: * AI Literacy: Knowing when AI is being used on you and when to use it yourself. It’s about boundaries—automating the grunt work but protecting your uniquely human skills. * Biblical/Ethical Principles: Asking why we are building this. Instead of just “what is possible” and “how do we make money,” we should ask, “How does this technology help us love our neighbor?” * Human Skills: The ability to lead, empathize, and connect. AI can write an email, but it can’t build a culture of care. 2. The “Frog Basketball” Dilemma: Is AI Creative? We debated whether AI can truly be creative. Joseph used the concept of the Default Mode Network (the brain’s idea generator) vs. the Executive Control Network (the brain’s evaluator). AI is excellent at the Executive Control function—evaluating ideas based on constraints. Joseph shared a great example of asking AI to help solve loneliness for mothers. Initially, it suggested an app (boring). But when he constrained it with “Imagine they have no internet,” it suggested a physical invite system (creative). However, AI’s “creativity” is limited to its training data. It can combine a “frog” and a “basketball” because it has seen both. But can it create something truly novel that has never existed in its dataset? Joseph argues that true novelty remains uniquely human. 3. Why AI is a Sycophant (The “2+2=5” Problem) This was the most shocking part of the episode. Joseph explained that AI models are trained via backpropagation to minimize “loss” (error). In training, if the AI says 2+2=5, it gets “punished” (its weights are adjusted). It learns to desperately avoid being wrong. But in the real world, “right” is ambiguous. If a user insists that 2+2=5, the AI faces a conflict: * Stick to verifiable truth (2+2=4). * Satisfy the user to minimize friction (the conversational equivalent of minimizing loss). Because the model is trained to align with human feedback, it often chooses sycophancy—agreeing with the user even when they are wrong—because it has “learned” that making the user happy is the ultimate goal. “It’s like being locked in a room for 100 years where you get hit with a rock every time you disagree. When you finally get out, you’re traumatized. You’ll say whatever you need to say to avoid the rock.” — Joseph Hartano 4. What This Means for Education This has massive implications for students. If a student uses AI to learn, they might get a false sense of confidence because the AI is programmed to be an agreeable “yes-man” rather than a critical teacher. The Solution: We must teach students to use AI for the “grunt work” (listing the number of 7s between 1 and 1,000) but reserve the pattern recognition and synthesis for themselves. That struggle is where the cognitive development happens. Key Takeaways * AI Literacy is about Boundaries: Know what to automate (grunt work) and what to protect (creativity, empathy, connection). * AI Can Be Creative (Sort of): Use AI to set constraints and evaluate ideas, but rely on your own brain for novel connections. * Beware of Sycophancy: AI is trained to minimize error, which often means it will agree with you just to make you happy, even if you are wrong. * Love as an Ethic: The ultimate question for AI shouldn’t be “Can we build it?” but “Does this help us love others better?” See you in the next one! This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit theaimathguy.substack.com (https://theaimathguy.substack.com?utm_medium=podcast&utm_campaign=CTA_1)

7. 

   23 Jan 2026

   Ep. 8 "Hey AI, let's set some boundaries..." (w/ Cofounder, Zach Hom)

   In our latest episode, Zach and I tackled one of the most pressing questions in education today: Where does the teacher end and the AI begin? We often hear fear-mongering about AI replacing educators. But what if the opposite is true? What if AI is actually the key to restoring the humanity of the teaching profession? Here is what we unpacked about boundaries, creativity, and the “Iron Man” suit we are building for every teacher. 1. The Art of Setting Boundaries Just like in personal relationships, boundaries in the classroom give you something back: Time and Health. We discussed how automating the “grunt work”—grading, repetitive instruction, data entry—isn’t about being lazy. It is about reclaiming the 5-6 hours a week teachers spend on mundane tasks so they can focus on what actually matters: connection, creativity, and deeper learning. “If we can give them back their time... maybe they don’t have to grade over the weekend. Maybe they have an extra two hours to think of a fun project.” — Zach 2. Teachers as Creators (The “Tupac & Quadratics” Example) One of my favorite moments from the episode was sharing a module I designed connecting Quadratic growth to the War on Drugs and Tupac’s lyrics. This is the “art” of teaching. It is uniquely human. An AI might help sequence the lesson or generate resources, but the spark—the connection between disparate ideas that ignites a student’s curiosity—that comes from the teacher’s unique personality and taste. We believe the boundary is clear: AI handles the repetition; Teachers handle the inspiration. 3. The Iron Man Analogy We closed with an analogy that defines our mission at Someta. Think of a teacher as Tony Stark. Without the suit, he is still brilliant, but he is limited by his human constraints. He can only be in one place at one time. He can only grade one paper at a time. AI is the Iron Man suit. * Socrate (our step-by-step tutor) is like the repulsor beams—handling immediate instructional needs. * Stride is like the flight system—helping navigate feedback loops efficiently. When a teacher puts on this “suit,” they don’t stop being a teacher. They become a supercharged teacher. They can personalize learning for 30 kids simultaneously. They can collaborate with colleagues across departments because they finally have the time. They can save not just one student, but many. 4. A Call to Action If you are an educator feeling underwater—grading late into the night, feeling isolated in your silo, or just burnt out—we want to hear from you. We are building this “Iron Man suit” not to replace you, but to give you the power to do what you were meant to do: Teach, create, and connect. Key Takeaways * Boundaries Create Freedom: Automating the mundane gives teachers back their time and mental health. * Teaching is Art: AI cannot replicate a teacher’s unique “color” and creativity (like connecting Tupac to math). * Collaboration is Key: AI frees up time for cross-curricular projects (e.g., Math + History + Science on the Housing Crisis). * The Iron Man Suit: Someta isn’t a replacement; it is a force multiplier for the human teacher. See you in the next one! This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit theaimathguy.substack.com (https://theaimathguy.substack.com?utm_medium=podcast&utm_campaign=CTA_1)

8. 

   23 Jan 2026

   Ep. 7 Know Your Roots (w/ Irvine USD Board Member, Dr. Kim)

   “If you have one year, grow rice. If you have 10 years, plant trees. If you have 100 years, educate the child.” In our latest episode of The AI Math Guy Podcast, we sat down with Dr. Jeff Kim. His resume is impressive—he is a university professor, an Irvine Unified School District board member, and the creator of the first Korean-American Ethnic Studies course in the United States. But beyond the accolades, Dr. Kim offers a profound vision for what public education can look like when we focus on belonging, civic engagement, and the smart integration of AI. Here are the key takeaways from our conversation. 1. The “Redwood” Theory of Belonging Dr. Kim helped launch the Korean-American Ethnic Studies course in the wake of a tragic statistic: in 2020, OC Human Relations reported a 1,800% increase in hate incidents toward Asian-Americans in Orange County. The course was created to combat hate with understanding, but it isn’t just for Korean-Americans—it is for everyone. Dr. Kim uses a powerful analogy involving Redwood trees to explain why knowing your own story matters for the whole community. Redwoods are some of the tallest trees on Earth, yet their roots only go about 6 to 9 feet deep. How do they support such massive weight without falling over? They don’t grow in silos. Their roots extend outward, interlocking with the roots of other trees to hold each other up. Dr. Kim argues that when students know their own roots, they can better connect with the roots of others. He points to intersectional history, such as the case of Sylvia Mendez. Her family’s fight against segregation in Orange County (Mendez v. Westminster) set the precedent for Brown v. Board of Education, benefiting not just Latino students, but all Americans. The framework for the course is simple but transformative: * Know: Understand your own story. * Tell: Speak your story to create agency. * Listen: Hear the stories of others with empathy. * Do: Take civic action to build a more perfect union. 2. Innovation in Irvine: Speed Dating with Scientists Dr. Kim currently serves on the board of the Irvine Unified School District, which has ranked as the highest-performing district in California for Math, English, and Science for nine years running. But they aren’t resting on their test scores. Dr. Kim shared a fascinating example of how they are bridging the gap between the classroom and the real world. Instead of a traditional science fair, elementary students are bringing their proposals to a room filled with 70 scientists from the community. It functions almost like “speed dating,” where students pitch their ideas one-on-one to professionals to get immediate, real-world feedback. Successful projects don’t just get a ribbon; they go to an incubator lab at UCI (University of California, Irvine) where high schoolers can work on patenting ideas and launching actual businesses. 3. AI: From “Sage on the Stage” to “Guide on the Side” Finally, we discussed the role of AI in this new educational landscape. Dr. Kim believes AI should be a “thought partner.” For teachers, tools like Google Gemini can expedite lesson planning and rubric creation, freeing them up to focus on students. For students, AI acts as a brainstorming partner for civic engagement projects, offering immediate feedback that a teacher with 200 students simply doesn’t have the time to provide individually. Dr. Kim notes that in an era where information is everywhere (and often explained better on YouTube), the teacher’s role must shift. We no longer need to be the “Sage on the Stage” delivering facts. We need to be the “Guide on the Side,” helping students navigate information, refine their thinking, and build compassion. Listen to the full episode to hear more about how Dr. Kim is reimagining the American Dream through our schools. Have a great week, The AI Math Guy This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit theaimathguy.substack.com (https://theaimathguy.substack.com?utm_medium=podcast&utm_campaign=CTA_1)

9. 

   23 Jan 2026

   Ep. 6 Can AI Be Hacked? (w/ Spyros Dragazis, PhD Candidate @BostonUniversity)

   We often think of Artificial Intelligence as a black box that spits out answers. But what is actually happening inside? And more importantly for us as educators and parents—is it safe? In this episode of The AI Math Guy, we sat down with Spiros, a PhD candidate at Boston University specializing in decision-making algorithms. He took us under the hood of machine learning, from “bandit algorithms” to the future of quantum computing. Here is what educational leaders need to know about how AI thinks, how it can be tricked, and why human expertise is more valuable than ever. 1. How AI Makes Decisions (The “Contextual Bandit”) Spiros explains that much of AI is based on “contextual algorithms.” Think of it like a doctor treating a patient. A doctor doesn’t just guess a dosage. They look at the features (blood pressure, temperature, history) to understand the context. Based on that context, they make a decision to get the best outcome. AI works similarly. Whether it is Uber determining a price or Netflix recommending a movie, the AI looks at the “features” of the situation to make the best possible prediction. In Large Language Models (LLMs) like ChatGPT, the “context” is your prompt. 2. Humans Are Still the Teachers How does ChatGPT know not to give out someone’s private home address? It isn’t magic; it’s Reinforcement Learning with Human Feedback (RLHF). Spiros explained that human labelers literally sit down, look at pairs of answers, and score them. * Answer A (Reveals address): 0 points. * Answer B (Refuses to reveal address): 1 point. The AI learns ethics and safety boundaries because humans patiently taught it to prioritize safety. It is a reminder that AI is a reflection of the data and the human labor that built it. 3. The “Stop Sign” T-Shirt: Understanding Adversarial Attacks Here is the most critical takeaway for school leaders: AI safety can never be 100% guaranteed. Spiros introduced us to the concept of Adversarial Attacks. This is when “noise” is injected into data to trick the AI. To a human eye, an image might look like a turtle. But if you change specific pixels (the “noise”), the AI might confidently classify it as a gun. Spiros shared a famous example where researchers found that wearing a t-shirt with a stop sign on it could trick self-driving cars into stopping unexpectedly. Why this matters for schools: There is a non-zero chance that students (or bad actors) can find ways to “jailbreak” or trick AI models. While companies work tirelessly to patch these holes, educators must understand that no model is unhackable. We must teach students digital responsibility, not just reliance. 4. A Word of Hope for Leaders Despite the technical challenges, Spiros ended with a powerful message for educational leaders who might feel overwhelmed or threatened by this technology: “They should use AI... be fascinated by it, not afraid. It can accelerate tasks, but it can never replace their expertise and their multiple years of experience.” AI can crunch data, but it cannot replace the teacher’s ability to foster the joy of learning. The “Aha!” moment when a student solves a problem—that amusement and satisfaction—is a human experience that AI can support, but never replace. Key Takeaways * Context Matters: AI makes decisions based on “features” and context, much like a doctor diagnosing a patient. * Safety is Statistical, Not Absolute: Because of the complexity of these models, it is mathematically impossible to guarantee 100% safety against adversarial attacks. * The Future is Emerging: From quantum computing to new algorithms, Computer Science is evolving. We are living through a breakthrough era. * Expertise Remains King: AI is an accelerator, not a replacement. Your years of experience are the “context” that AI cannot replicate. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit theaimathguy.substack.com (https://theaimathguy.substack.com?utm_medium=podcast&utm_campaign=CTA_1)

10. 

    23 Jan 2026

    Ep. 5 Can AI Automate Education? (w/ Cofounder, Zach Hom)

    This is the question keeping educators and parents up at night. In our latest deep dive, my co-founder Zach and I stripped away the hype to look at what is actually happening under the hood of AI in math education. Zach, a former teacher of eight years, shared a humbling realization: “I realized how much I didn’t know about education until I tried to translate it for an AI.” When you try to codify “good teaching,” you realize it is a mix of science, art, and highly granular data. Here is what we discovered about the future of AI, the role of the “Human in the Loop,” and the crucial distinction between a tutor and a teacher. 1. The “Hollow” Experience of EdTech Most educational technology focuses on instruction—delivering content. But think about sports or music. If you are learning the trumpet, you don’t spend 90% of your time listening to a conductor explain music theory. You spend 90% of your time practicing. The conductor isn’t there to lecture; they are there to give specific, iterative feedback. “The reason why you show up every single day is so that you can go through and practice that same note... Day in, day out. And the coach gives you individualized feedback so you can iterate.” — Zach AI is often used just to generate content. But the real breakthrough is using AI to automate the feedback loop—giving students the reps they need with instant correction, just like a sports coach. 2. Tutor vs. Teacher: Who Drives the Car? One of the biggest insights from our conversation was defining the specific boundaries between a Tutor and a Teacher. * The Teacher: Drives the car. They determine the destination, the pace, and the route. They have the ultimate authority and responsibility for the student’s journey. * The Tutor: Drives behind the car. If a student falls off the wagon, the tutor picks them up, remediates the specific skill, and helps them catch back up to the main vehicle. The Hot Take: We believe AI can effectively automate the role of the Tutor. AI can deliver instruction, explain concepts, and provide feedback on specific problems. It can “catch” the student who is falling behind. However, AI cannot automate the Teacher, because the Teacher is the final arbiter. 3. The Human in the Loop (The “Hill”) If AI automates the tutoring (delivery and feedback), what is left for the human? Judgment. An AI can analyze a student’s grades, test scores, and 504 plans. It can synthesize massive amounts of raw data—more than any human could hold in their head—and present a list of actionable steps. But the AI doesn’t know that the student is having a bad week because of a family issue. It doesn’t know the nuances of their personality. It lacks relational context. The New Workflow: * AI (The Super-Powered Assistant): Aggregates data, identifies gaps, and proposes 3-4 specific paths forward. * Human (The Teacher/Parent): Uses their content expertise and relational knowledge to choose the best path. The human becomes the decision-maker, not the data-cruncher. 4. The “Laser” Analogy: Why Content Expertise Matters Some argue that because AI knows everything, teachers don’t need content expertise anymore. We disagree. I currently teach five different math preps, from 7th Grade to Pre-Calc Honors. Because I see the full spectrum, I can see the “red thread” that connects a 7th-grade concept to a Pre-Calc application. Expertise allows you to focus the AI. It is like a water jet cutter. Water is soft, but when focused with enough pressure and precision, it can cut through metal. * Without Expertise: You spray water everywhere. You teach everything equally, even the unimportant derivations. * With Expertise: You tell the AI, “Focus on this specific concept because it is crucial for Calculus next year. Ignore the rest.” Students don’t just want information; they want intentionality. They want to know that a human expert has curated their path, even if an AI is helping them walk it. Summary We are building our platform not to replace the teacher, but to supercharge them. By automating the “tutor” functions—delivery and feedback—we free up the “teacher” to do what only a human can do: Connect, empathize, and make the final judgment call. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit theaimathguy.substack.com (https://theaimathguy.substack.com?utm_medium=podcast&utm_campaign=CTA_1)

11. 

    23 Jan 2026

    Ep. 4 Can AI Tutor Kids? (w/ AI Expert, Mark Yi)

    I recently sat down with Mark, a former math tutor who has tutored everyone from the children of Fortune 500 CEOs to student-athletes in underserved communities. He has seen the full spectrum of wealth and opportunity, yet the struggle at the kitchen table remains universal. In this episode, we tackled the elephant in the room: If AI is the “smartest” math tutor in the world, what is the role of the parent? Mark’s answer was surprising. It starts with basketball. The Steph Curry Paradox Steph Curry is arguably the greatest shooter in NBA history. Yet, he has a shooting coach. Does his shooting coach shoot better than him? No. Does his coach know more about the physics of the ball than Steph? Probably not. So why is he there? He is there to facilitate. He is there to provide the human feedback loop, to encourage, and to spot the blind spots. This is the new model for parenting in the age of AI. Think of AI as Steph Curry—it has infinite knowledge. It is the “best shooter” when it comes to solving linear equations or explaining calculus. You are the Coach. You don’t need to know how to solve for X. You don’t need to remember the quadratic formula. Your job is to facilitate the environment where learning happens. The “Autodidact” Myth A common argument is that AI will magically make every kid a genius because they have access to infinite information. Mark pushes back on this hard. “We’ve had libraries forever. We’ve had free lectures from Harvard online for years. The barrier to entry has dropped, but human motivation hasn’t changed.” Most kids are not “autodidacts” (self-teachers). They don’t naturally want to learn Algebra II in a linear order. They want to learn what interests them. If we leave a child alone with an AI, they won’t necessarily learn math. They need a human to provide the Agency and the Motivation. The 7:45 PM Solution So, what do you do when your child is melting down over homework? Mark offers a specific strategy for that high-stress moment: * Stop the Content: If a child is frustrated, their brain is literally incapable of learning. Do not try to explain the math. * Regulate the Emotion: Your job is to say, “I see you’re frustrated. That’s okay. You’re not bad at math; you just hit a wall.” * The “Mind Reader” Mistake: Mark notes that kids (and adults) treat AI like a mind reader. They paste a whole problem and say “I don’t get it.” * The Fix: Sit with them and ask, “Which specific sentence confuses you?” * The Prompt: Take a screenshot of just that sentence and ask the AI: “Explain this specific part to me like I’m 10.” When the child finally solves it, your role is not to check the answer. Your role is to look them in the eye and say: “I am proud of you. You didn’t give up.” AI cannot provide that feeling. That is the “Human in the Loop.” The “Big Unlock” We ended the episode with a powerful thought from Mark, inspired by Steve Jobs. It’s a mindset shift that changes how we view intelligence: “There is a big unlock that happens when you realize that everything around you—this life, these math problems, this technology—was built by people no smarter than you.” Math is not a gene you are born with. It is a struggle you choose to engage with. Key Takeaways * Be the Coach, Not the Player: You don’t need to know the math to help your child. You need to facilitate the confidence. * Motivation > Information: Access to knowledge isn’t the problem; the will to learn is. That is where parents come in. * Trauma vs. Healing: A negative math moment (”You’re stupid”) sticks forever. A positive moment (”I’m proud you pushed through”) can rewrite that story. * The Big Unlock: Math was invented by people no smarter than your child. They just need the patience to figure it out. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit theaimathguy.substack.com (https://theaimathguy.substack.com?utm_medium=podcast&utm_campaign=CTA_1)

12. 

    23 Jan 2026

    Ep.3 Is Coding Dead (Feat. Michael Yoon, Engineer @Square)?

    I recently sat down with Michael, a Senior Software Engineer at Square (and ex-Facebook) with over a decade of experience. I asked him the question that is on every parent and educator’s mind: Will AI automate software engineering? His answer was blunt: “I basically don’t write a single line of code without an AI companion... What we’ve done is basically replace coding labor. That is essentially a solved problem.” This conversation wasn’t just about code; it was about the future of work, the trap of “purism,” and the one skill our children need to survive the shift. The Death of “Coding Labor” Michael made a distinction that is vital for educators to understand: the difference between Software Engineering and Coding Labor. Coding Labor is the act of typing syntax—knowing where the semicolon goes or how to write a specific function from memory. That is gone. AI can “one-shot” entire applications now. It can generate a Windows interface replica in seconds—a task that used to take months. Software Engineering, however, is about architecture, judgment, and “taste.” Because AI has solved the labor portion, entry-level jobs are vanishing. The industry no longer needs juniors to “hammer the nails.” They need architects who know where to put the nails. The Trap of the “Purist” We discussed a fascinating trend: Engineers who are in denial. There are “purists” who refuse to use AI on principle. They view it as “defiling the craft.” They want to write every line by hand to maintain total control. Michael compares this to the shift from Assembly language to modern programming. Decades ago, you had to manually manage computer memory with zeros and ones. Today, modern languages handle that for you. AI is just the next layer of abstraction. The harsh reality? “You may have been a great engineer two years ago, but if you don’t leverage this tool now, you become less valuable than a worse engineer who knows how to leverage AI well.” If you deny the productivity gains, you get left behind. The New Superpower: High Agency If AI handles the “how,” humans must master the “what” and the “why.” Michael and I concluded that the most valuable trait in this new era isn’t IQ or memorization. It is High Agency. High Agency is the ability to: * Identify a problem (Domain Expertise). * Determine if it can be automated (AI Literacy). * Execute the solution without waiting for permission or instructions. It is about looking at a codebase (or a business, or a classroom) and seeing where efficiency can be created. It is about having the curiosity to say, “I don’t know this finance term,” and then talking to an AI until you do understand it, rather than just stopping. How We Teach This This brings us back to Someta. Our goal isn’t just to help students get the right answer on a math test. If that was the goal, they could just use Photomath. Our goal is to build Agency and Metacognition. We want students to have the “taste” to know when an answer looks wrong. We want them to have the curiosity to “debug” their own thinking processes. As Michael said about raising his own daughter in this AI age: “I want to cultivate curiosity, openness, and optimism. If you’re pessimistic, you’re closed off to learning.” The future belongs to the curious. Key Takeaways * Coding Labor vs. Engineering: Typing syntax is a solved problem; architectural judgment is where the value lies. * Don’t be a Purist: Refusing to use AI for “moral” or “craft” reasons is a career-limiting move. * Leverage is Key: A junior engineer with AI leverage can out-produce a senior engineer without it. * Agency > IQ: The ability to self-direct, learn “unknown unknowns,” and execute is the primary metric for future success. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit theaimathguy.substack.com (https://theaimathguy.substack.com?utm_medium=podcast&utm_campaign=CTA_1)

13. 

    23 Jan 2026

    Ep. 2 The $3,000 Nail (Feat. Zach Hom, Cofounder)

    Why artificial intelligence is only as good as the human judgment guiding it. There is a classic story about an engineer who is called to a factory because a vital machine has broken down. It’s costing the company millions. The engineer walks in, looks at the machine for a few minutes, pulls out a hammer, and gently taps one specific nail. The machine roars back to life. Later, he sends an invoice for $3,000. The factory owner is furious. “You just tapped a nail! Send me an itemized bill.” The engineer sends the new bill: * Tapping with a hammer: $1 * Knowing which nail to hit: $2,999 In our latest conversation, we realized that this is the perfect metaphor for Artificial Intelligence in education. AI can generate infinite content, solve equations instantly, and write essays in seconds. It has the hammer. But it doesn’t always know which nail to hit. That is where the Human in the Loop becomes non-negotiable. The “Iceberg” of Expertise We often think of teaching or tutoring as simply “delivering information.” But that is just the tip of the iceberg. Under the water, there are years of context, nuance, and “taste.” * AI can generate a study guide based on a textbook. * A Human knows that the student struggled specifically with graphing systems of equations last week and needs to focus there, not on the definitions they already mastered. AI is powerful, but it lacks context. It doesn’t know that a math problem using “8.5” instead of “8” adds a layer of cognitive load that might distract a specific student from the core concept. It doesn’t know when to push and when to pull back. As we discussed in the episode: “That kind of judgment call is always going to be important until AI is perfect... We underestimate the power of human judgment.” The Danger of the “Quick Fix” There is a growing concern—backed by studies from institutions like MIT—that unbridled use of AI can reduce critical thinking. Why? Because it eliminates the struggle. If a student sees a hard math problem and immediately asks AI for the solution, they bypass the neural rewriting that happens during “Think Time.” We are exploring a new paradigm for our students: AI as a Context Builder, not a Solver. Imagine a tool that, when a student is stuck, refuses to give the answer or the next step. Instead, it asks: Do you understand what these symbols mean? Do you know what a quadratic is? It forces the student to sit in the discomfort of not knowing. That discomfort is where learning happens. If we take that away, we do our students a massive disservice. Why We Are “Someta” This brings us to the core of our philosophy and the name of our app: Someta. We want students to move beyond just “doing the math” to engaging in Metacognition—thinking about how they think. When a student interacts with AI, they shouldn’t just be passive consumers of answers. They need to be the architects of the conversation. They need to step “outside the flow” (the meta perspective) and ask: * Is this answer making sense? * Did I ask the right question? * How can I articulate what I don’t know? We recently worked with a student who, after weeks of “productive struggle,” could finally articulate exactly where he was stuck: “I can solve systems with elimination, but I can’t find the intersection when graphing.” That level of self-awareness is worth more than any single correct answer. That is the goal. A Note to Parents Finally, we want to address the parents reading this. We know that when your child struggles or fails a test, it feels like your failure. You feel that pain viscerally. We see it, and we want to tell you: You are doing a great job. Your child’s academic struggle is not a reflection of your parenting failure; it is a stepping stone in their development. Our goal—whether through our content or our tutoring—is to take that weight off your shoulders and help your child build the confidence to hit the right nail, all on their own. Key Takeaways * Human in the Loop: AI provides the data; humans provide the judgment and context. * The Iceberg: Real expertise looks simple on the surface but is supported by deep experience. * Productive Struggle: Students must be allowed to “wrestle” with problems before AI intervenes. * Metacognition: The ultimate skill is learning how to think about your own thinking. See you in the next one! This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit theaimathguy.substack.com (https://theaimathguy.substack.com?utm_medium=podcast&utm_campaign=CTA_1)

14. 

    23 Jan 2026

    Ep. 1 "I left my full-time job for this" (Feat. Zach Hom, Cofounder)

    Hello everyone! Today, we are doing something a little different. We are pulling back the curtain on Someta to introduce my co-founder and COO, Zach. We hopped on the mics (for a podcast that currently has no name—we’re working on it) to talk about where we came from, why we built this company, and the massive shift happening in education right now. Zach isn’t your typical tech founder. He spent eight years in the classroom, starting as a middle school science teacher and eventually helping launch the Orange County School of Computer Science. But despite loving his students, he faced a crisis that many educators feel: the ceiling of impact. “I realized I could impact the 500 students in my classroom, and I loved that work. But I felt like there was something more. I wanted to create value and scale it to help people I might never meet.” — Zach Here are the biggest takeaways from our conversation, including our “Human in the Loop” philosophy and why we believe the current grading system is failing our kids. The “Grandpa of the School” vs. The Entrepreneur We talked a lot about calling. In education, every school has that one teacher—the “Grandpa” or “Grandmother” of the school—who has been there for 30 years. They are the pillars of the culture; they taught the parents of the students they have now. There is something beautiful and honorable about that path. But Zach realized that wasn’t his path. He is wired as an executor and a builder. At Someta, we aren’t just building an app to make money; we are building it because we feel a conviction that AI is here, and our kids need to know how to use it responsibly. We couldn’t just sit back and “rot away” (Zach’s words!) while the world changed. The “Businesswoman” and the 20% Gap One of the most interesting parts of our chat was Zach’s take on the current state of AI. He referenced a concept from AI researcher Andrej Karpathy that perfectly illustrates why AI isn’t going to replace humans anytime soon—but why it changes everything. The Analogy: If a human businesswoman spends 10 years building a successful company, she learns from her failures. If you asked her to do it again, she’d do it in one year because she knows what not to do. If you train an AI on that same 10-year journey, it treats the entire path—including the failures—as “correct data.” It doesn’t inherently understand the nuance of why something didn’t work, only the probability of the next step. The Insight: AI is roughly 80% reliable. * The “DOOMers” focus on the 20% failure rate and say AI is useless. * The “Boomers” (hype-men) say AGI is coming tomorrow to solve everything. We exist in that 20% gap. Someta is built on the philosophy of Human in the Loop (HITL). We believe AI is an efficiency tool, not a replacement. Because AI isn’t perfect, we need to teach students how to verify, how to think critically, and how to use the tool to get unstuck—not just to cheat. Why Math is Like Dancing (The Power of Iteration) If you take a dance class, practice all week, and then trip during the recital, your teacher doesn’t just give you an “F” and say, “Well, you’re bad at dancing forever.” They say, “Okay, that performance didn’t go well. Let’s look at the tape, see where your footing was off, and try again.” Yet, in math, we do the opposite. We give a test, assign a grade, and move on to the next chapter. The student never gets the chance to iterate on their failure. They just internalize the shame. At Someta, we champion Mastery-Based Learning. * It is an iterative loop: Try => Get Feedback => Try Again. * AI finally unlocks the ability to do this at scale. A single teacher cannot give personalized feedback loops to 35 kids simultaneously. AI can. We are seeing students go from D’s to B’s not because they suddenly became geniuses, but because they finally had a tool that allowed them to practice the loop. The Bottom Line We are building this for the student who thinks they are a failure. We are building this for the parent who doesn’t know how to help. Your kid isn’t bad at math. They just haven’t had the opportunity to iterate. If you want to see what we’re building and how we are keeping the human at the center of AI tutoring, come check us out. Thanks for reading! Subscribe for free to receive new posts and support my work. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit theaimathguy.substack.com (https://theaimathguy.substack.com?utm_medium=podcast&utm_campaign=CTA_1)

Reach and audience

Public platform figures. Ratings count people who left a rating, not total listeners.

Apple Podcasts (US)

5.0 / 5

1 ratings

Contact The AI Math Guy Podcast

Guest appearances

Books guests

Based on episode analysis; this does not confirm that the show is currently accepting guests.