The Operating System Between You and AI: A Conversation with M21 Gavin Lee

Gavin (Chia Ta Lee), MDA 2021, is the first Taiwanese admitted to Minerva University's Master of Decision Analysis program, co-author of Taiwan's bestselling book on Habits of Mind, and founder of Worldviews Academy. He has trained over 5,000 students across Taiwan's top EMBA, universities and corporations. In this interview, he shares how a career of high-stakes crises led him to Minerva — and why he believes Habits of Mind are the most important skill set of the AI era.

Tell us about yourself. What brought you to Minerva?

I was a lawyer. Chief Legal Officer of a tech company with over NT$2 billion in capital. Then the world fell apart. International markets collapsed. We ran out of cash. Couldn't pay next week's bank interest on a syndicated loan. Over 200 employees' futures hung on one negotiation.

The banker across the table had a stated position. But that wasn't what mattered. What mattered was his real concern — he was terrified that all monitoring responsibility for the restructured loan would fall on his bank alone. Nobody else at the table saw it.

I did. Because I wasn't listening to what he said. I was analyzing why he was saying it. That one insight saved the company. And it taught me something I couldn't unlearn: the quality of your thinking determines the quality of your outcome. Not your credentials. Not your connections. Your thinking.

This pattern repeated in every crisis I faced. Every corporate meltdown, every legal battle, every impossible negotiation. The people who survived weren't the smartest in the room. They were the ones who thought most clearly under pressure. I became obsessed with one question: can decision-making be taught?

That question led me to Minerva. The first Taiwanese admitted to the Master of Decision Analysis program. After graduating, I founded Worldviews Academy. Five years later, we've trained over 5,000 students — across Taiwan's top EMBA, universities and some of the country's largest corporations. Every classroom runs on active learning. No lectures. No passive note-taking.

Because I learned the hard way: knowing isn't the same as doing. And thinking is something you practice, not something you're born with.

You co-authored a book on Habits of Mind that became a bestseller. What made it resonate so deeply?

A comment from one of our readers provides the best interpretation: "Habits of Mind is the best index I've ever seen for everything I've learned." That sentence explains why it worked.

The book is called The Habits of Mind the World's Top Talents Are Racing to Learn (全球人才搶著學！密涅瓦的思考習慣訓練). It hit the top 100 business books on Books.com.tw — Taiwan's largest online bookstore — in 2022. Three of us wrote it. A lawyer, an ER doctor, an international sales director(MDA 2021 and 2022). We didn't write theory. We wrote what happened when we applied Habits of Mind to real crises — in the courtroom, the emergency room, the boardroom.

That's what made it different. There are some books about what Minerva teaches. Ours is the only one about actually using it from students' perspectives. After the book came out, something unexpected happened. Minerva alumni from around the world wrote to us that the book was the reason they decided to enroll.

That told me we'd hit something real. Not because we were brilliant writers. Because practitioners sharing actual professional crises will always land harder than academics theorizing from a distance. The book didn't explain Habits of Mind. It proved them.

What's the most important lesson you took from Minerva?

A friend of mine commented on Minerva's pedagogy after reading our book. His words resonated with me the most: "Minerva uses the newest methods to learn the most classic thinking." That sentence unlocked the entire system. Because Minerva's Habits of Mind aren't random. They run on two tracks. And once you see the tracks, you see everything.

Track one: Formal Analysis. This is deductive logic. You take a concrete, messy, real-world problem and abstract it into a universal model. You stop seeing symptoms. You start seeing patterns. Here's a real example from my own life. My twins started kindergarten. Their teacher took hundreds of photos every week with a DSLR and uploaded them to a shared Google Drive folder. Hundreds of photos. Mixed with every kid in the class. No labels. No sorting. A growing heap of files every single week.

The instinct is to solve the symptom. Download manually. Sort one by one. Spend every weekend as a human photo filter. But formal analysis demands something different. It demands that you stop looking at the mess and start looking at the structure. So I built a model. Any problem — this one or any other — can be decomposed into four components: the goal state, the constraints you cannot change, the obstacles you can remove, and the solution space that remains.

Goal: my children's photos automatically identified, classified by activity, and stored in Google Photos. No manual steps. Constraints — the walls that cannot move: I cannot ask the teacher to change how she shares. I cannot request she tag my kids specifically — that would make me the helicopter parent no teacher wants to deal with. And I refuse to pay. Zero dollars. Non-negotiable.

Obstacles — the walls that can be demolished: there is no mechanism to automatically identify my children in a pile of two hundred photos. There is no system to sort and upload. Both are solvable. Once the model was clear, the solution was obvious. A Python script. Google's Gemini Flash API for face recognition — free. Automatic classification, automatic upload, automatic cleanup. Total cost: zero. Time per batch: seconds. The original problem felt overwhelming — an emotional mess of too many photos and not enough time. The model made it trivial. Not because the tools were clever. Because the formal structure stripped away the noise and left only the variables that mattered.

That is formal analysis. You take any concrete, messy, emotional situation and abstract it into a universal model: goal, constraints, obstacles, solution space. Once you see this structure in one problem, you start seeing it in every problem you encounter.

Track two: Empirical Analysis. This is inductive logic. Observe. Hypothesize. Test. Validate. Patterns aren't for admiring. They're for testing and acting on.

Formal analysis builds the map. Empirical analysis walks the territory. Together, these two tracks are the engine behind all Habits of Mind. Every HC — whether it's about evaluating evidence, identifying assumptions, or weighing alternatives — traces back to one of these two modes of reasoning.

I tell my students: "These two thinking modes are not only the core design of Minerva's education. They are the tools for standing undefeated in the AI era." Here is why. AI can now generate ten models in seconds and propose hypotheses you never considered. It is brilliant at producing options. But it cannot know what actually matters in your situation — the two hundred employees whose futures depend on your next move, the cultural constraint no outsider would see, the risk only you can feel.

The stronger AI gets at generating possibilities, the more critical your ability to evaluate them becomes. Formal and empirical analysis are how you evaluate — and how you know whether AI is taking you closer to the answer or further away. That is your job. Not to compete with AI. To orchestrate it. And these two tracks are how you do it.

How has Minerva's training shifted the way you make decisions — and why is it especially critical today?

The truth about AI is "orchestrating," not "prompting." Most people open ChatGPT and have a conversation. They ask a question. They get an answer. They think that's AI. It's not. That's typing with extra steps. The truly skilled people — before they open AI — already have an operating system loaded in their mind. They know what to ask, why to ask it, and how to evaluate the answer before it appears. Habits of Mind is that operating system.

McKinsey published a study that the top 6% of AI users achieve 10.3x productivity gains. The next 33% get 3.7x. And the bottom 61% — people who invested in AI tools — got nothing.

Zero.

What separates the top 6% from everyone else? Not better tools. Not more prompts. A better mental operating system. Without HC, AI gives you what I call "correct garbage." Technically accurate outputs that are strategically useless. The grammar is perfect. The logic is airtight. And it completely misses the point.

With HC, everything changes. AI becomes a thinking partner that examines problems from multiple angles, surfaces blind spots you didn't know you had, and maps chain reactions three moves ahead. HC is the interface between you and AI. It's the layer that turns raw computing power into actual intelligence. Think of it this way: AI is an engine. A powerful, tireless, brilliant engine. But an engine without a driver goes nowhere. HC is the driver.

And right now, most people are sitting in front of the most powerful engine ever built — with no idea how to steer.

What advice would you give to current Minerva students and alumni?

Three steps. In this order.

First: rewrite HCs for your own context — build a list that is truly yours.

Be honest with yourself. Out of all HCs, most people truly master five or six. The rest are things you've heard of, maybe studied, but never internalized. The habits only work when you consciously, deliberately apply them. Start by identifying which ones you actually use under pressure — then rewrite them for your real scenarios. That's your real HC list.

Let me show you how far this idea can travel.

I joined a community called "The Wayfinder" — a group dedicated to reviving traditional Austronesian celestial navigation. We practice traditional sailing. No phones. No GPS. No smartwatches. We learn to read the stars, the swells, and the wind the way Pacific voyagers crossed thousands of miles of open ocean centuries ago.

Papa Mau Piailug — one of only four master celestial navigators remaining in the world — once said: "You cannot bring paper into the ocean. When problems arise, a good navigator must commit fully and observe the changes." That is far transfer. Taking a Habit of Mind learned in a classroom and deploying it where there are no classrooms.

Out on the water, I wasn't learning astronomy. I was practicing asking the right problem under extreme conditions. What is my current position? What reference points can I find in the sky? How do I track time without a clock? How do I coordinate a crew that can barely see each other? Each of these is an obstacle to be analyzed. The binding constraint — no technology, no instruments — cannot be changed. Everything else can be worked.

The Wayfinder is not a metaphor for Habits of Mind. It is Habits of Mind. Stripped of slides, stripped of case studies, stripped of everything except you, the sky, and the quality of your thinking.

That is what rewriting HCs for your own context means. Not reviewing a textbook. Taking the same thinking framework and deploying it in your world — whether that world is a boardroom, an emergency room, or the open ocean at night.

Second: make it your organization's common language.

CE Tsao — Managing Director at Capital Group, one of the world's largest investment firms — shared how Habits of Mind transformed her sales team's conversations after taking my class. Her team members would come to her and say: "CE, this deal is impossible. The client's budget isn't enough."

That sounds like a constraint. Immovable. End of discussion. But CE would ask one question: "Who told you the budget isn't enough?"

Usually, it was a project manager.

"Can we reframe this? If you're talking to a project manager and the budget is X, what happens if we bring the conversation to his boss? Maybe the budget isn't X anymore. Maybe it's three X."

The key: CE didn't have to challenge her team member's effort. She didn't say, "Are you sure you tried hard enough?" She didn't question their commitment. Instead, she used a shared language — the language of constraints versus obstacles — to help them see the problem differently. The budget limit wasn't a wall. It was a door that opened at a higher level.

Her team knew exactly what CE could do: step in, have a face-to-face conversation with the decision-maker, and unlock budget that the project manager never had authority to offer.

Same meeting. Same client. Same product. Different thinking. Because the entire team shared a common Habits of Mind vocabulary. When teams share a thinking language, every conversation gets sharper. Disagreements become productive. Decisions improve.

Third: make it your AI operating system.

Feed your HC list to AI as the thinking foundation before any task. Not as decoration. As architecture. I discovered that Minerva's Habits of Mind can become the engine for commanding AI — directly helping you build AI into your personal super-brain.

But here is the honest truth: Minerva's course materials and student handbook alone are not enough. You need to conduct your own deep research on each HC you want to use. Fine-tune the definitions. Rebuild them into a custom knowledge base designed specifically for AI consumption. And layer in your domain expertise — the professional context that makes each HC actionable in your field. Only then does HC become a true AI operating system — not a concept you've studied, but architecture you've built.

Who should read this book, and what's the one thing you want them to walk away with?

Charlie Munger said: "You've got to have multiple models — because if you just have one or two that you're using, the nature of human psychology is such that you'll torture reality so that it fits your models." He built his career on that principle. Minerva's Habits of Mind are the most rigorous, teachable, testable version of that idea I've ever encountered.

But let me tell you a story about chess.

In 1997, IBM's Deep Blue beat Garry Kasparov. The greatest chess player in history, defeated by a machine. The headlines screamed that human chess was dead. The opposite happened.

Before computers entered chess, the world produced roughly 4.68 new Grandmasters per year. After computers? 53.6 new Grandmasters per year. An eleven-fold increase. Before 2010, only 4 players under age 14 had ever earned the Grandmaster title. After 2010? Twelve players under 14. The tools built to beat humans became the tutors that elevated them.

Kasparov himself saw it first. In 1998 — one year after his defeat — he invented "centaur chess." Human-AI teams. These centaur teams didn't just beat grandmasters. They beat supercomputers playing alone. Read that again. A human with a machine beat the machine by itself.

Kasparov said: "The solution isn't less technology, but better humans." And: "As our machines become more adaptable, more powerful, we do as well." This is the parallel that keeps me up at night.

HC is your chess engine. The more powerful AI becomes, the more powerful you become — if you have the operating system to direct it. Without that OS, you're Kasparov in 1997, losing to a machine. With it, you're the centaur — unbeatable. Munger's mental models. Minerva's Habits of Mind. Same idea, different packaging. Everyone needs a thinking toolkit. Not more information. A toolkit.

This book is that toolkit.

What are you working on now?

I'm building the bridge between Habits of Mind and AI. In my previous course "AI Era Thinking Guide" (AI時代思考指南) — over 3,000 enrollments — and in my newest course "Building Your AI Super Brain," I've proposed a completely new AI command system. HC isn't just a thinking framework anymore. It's the operating system for the entire human-machine interface.

Four commands. Command AI how to think — load the HC framework before you ask anything. Command AI how to think professionally — layer in your domain so it stops being generic. Command AI how to think as a team — let it challenge your blind spots the way a Minerva classmate would. Command AI how to grow your thinking — every output sharpens your HC for the next problem.

Most people give AI a task. I give it a thinking architecture. The difference is everything. Tools change generations. Thinking never expires.

The stronger AI gets, the stronger you get. But only if you have the operating system installed.

Where can people find the book and your work?

The book is available at all major Taiwan bookstores — Eslite, Books.com.tw, and other online retailers. The sequel is expected later this year, focusing on how to combine Habits of Mind into AI skill packages that solve real-world complex problems.

For corporate training, find us at Worldviews Academy (世界觀學院).

‍