Career change · August 2026

I Stopped Trying to Figure Out My Next Career

I left BCG without a clear next step. A few months of small experiments gave me better evidence than soul-searching did.

This piece was originally written in Chinese. The English version has been adapted for natural reading while preserving the original meaning.

Thinking Wasn’t Getting Me Anywhere

After leaving BCG, I spent months trying to think my way into a new career. I kept asking: What should I do next? Who was I supposed to become?

I got nowhere. Then I read Working Identity by Herminia Ibarra. Her advice was practical: when you do not know what comes next, test and learn. Try small experiments and let the results change your direction.

Doing something gave me new evidence. That became the most useful rule in my career change.

I Built Eight Products in Three Months

I started learning how to build with AI agents in February. I was not planning to teach. I simply wanted to make things.

Over the next three months, I built eight products, including climate tools, workplace tools, personal apps, and web art. I shared each one with friends and put my learning path, notes, courses, and examples on GitHub.

Only later did I learn the terms learning in public and building in public. At the time, I was following a natural rhythm: learn, make, document, explain.

A Friend Asked Me to Teach Him AI

One day, a friend asked if I could teach him how to use AI. I invited him and another friend to a café that afternoon. We started from zero and built something together.

More people began asking. Some returned with new projects. I realized that I liked teaching AI and that it could become real work.

I had done a version of this work at BCG: learn a new climate topic quickly, find the central question, simplify the complexity, and explain it to people with very different backgrounds. The subject and setting changed. The way I worked did not.

Each Experiment Answered One Small Question

Making products showed me that I like creating. Sharing them told other people what I was exploring. That first class in a café showed me how much I enjoy helping someone make an idea real.

The experiments did not tell me what I should be doing five years from now. They told me what to try next.

I’m Leaving the Answer Open

If I went through another career change, I would spend less time looking for a new identity in my head. I would start sooner with a few small experiments: build a product, teach one person, take on one real problem.

Then I would look at the evidence. What do people keep asking me to help with? Which old strengths keep resurfacing? Do I want to keep doing this?

I did not think my way into this work. I found it by doing it.