A mountain valley and lake.

I uncover the structure beneath a problem and translate that understanding into a clear system for action.

· Gorges du Verdon, France

Personal index / Shuwei Chang

Logic Architect · Systems Thinker · AI-assisted Problem Solver

Notes, systems, and experiments by Shuwei Chang.

Contact

ABOUT

My way of working is to break complex things down and uncover the structures that make them work.

I’m interested in a wide range of questions: mathematics, AI, business, writing, and the practical problems that arise in everyday life. They may seem unrelated, but I tend to approach them in much the same way:

Why does this work the way it does?
What is surface detail, and what really matters?
Could taking the problem apart differently reveal a clearer structure?

Once I find that structure, I try to turn it into something that works in practice—a method, a system, a tool, or an experiment.

This website is an index of that work.

It brings together my thinking, ongoing projects, and the process through which an initial question gradually takes shape as a system.

AI is a tool I use extensively. It allows one person to explore, test, and build things that once would have required a larger team.

But I don’t hand my thinking over to AI.

I remain responsible for defining the problem, deciding what is worth doing, recognizing when an answer isn’t good enough, and making the final judgment.

MATH SYSTEM

A mathematics problem is more than an answer and a solution. It is a structure of knowledge that can be taken apart, compared, and reorganized.

I’m building a system around past mathematics questions from Taiwan’s major exams, connecting the problems, concepts, methods, and reasoning behind their solutions.

The project also explores a question:

Once a large collection of mathematics problems has been structured, can AI go beyond solving individual questions to truly understand the relationships between them—and help students learn how to think?

The question bank, detailed solutions, review process, AI problem-solving, and learning tools have all grown out of exploring this question.

NU SKIN NETWORK MARKETING

I started building my Nu Skin business in 2006.

At the time, I already ran my own tutoring center, and the business was doing well. But I began to realize something: earning more, having more time, and living a better life don’t necessarily go hand in hand.

That led me to look for another way of working.

What I learned here went beyond sales or building an organization. It was about working with people, developing a way of working that others could replicate, and figuring out how a business could fit into the life I wanted to lead.

I’ve continued this work since 2006.

Twenty years later, I’m looking at it again, with a different question:

As AI changes how people relate to information and work, which parts of network marketing still need to be done by people? And which could be redesigned?

THINK

Reading is a constant part of my life. I read more than 300 books a year.

But what interests me most is what stays with me afterward.

A book, a passage, or something that happens will sometimes make me pause and reconsider a question I thought I already understood. I’ve made a habit of writing down those moments.

Some are just a sentence. Others follow a short line of reasoning or capture an observation about work, life, business, human nature, and how the world works.

I call this collection, built over many years:

THINK.

These are neither book summaries nor fully developed essays.

They are an ongoing record of my thinking—what I noticed, what I questioned, and how I understood the world at different points in time.

Now, these words live here too, alongside the moments that first gave rise to them.