To be continued.
Tech was the start, law was a detour, marketing research is the craft
Help first, don't deceive, good product fair price
Automation is not harassment, simplicity is aesthetic
Self, others, nature — the three nets you live in
Not compromise — judge at the intersection
To those who should not be here
This story is far from over.
The romantic's premise: an ideal is an ideal precisely because it can never be reached. The moment you arrive, it stops being an ideal.
"Knowing it cannot be done, and doing it anyway" — Confucius, Qu Yuan, Cui Jian. None of them won. And none of them lost. It is not idealism. It is a stance: you do not calculate whether you can win before you move. You check whether the stance is right, and then you move.
But I reject the label "idealist."
Idealism is an insult disguised as a compliment. It means you fail in reality. What you do is admirable but useless — because you didn't win. It is a hat other people put on you, not a choice you made for yourself.
The contradiction: every choice I make — help first, knowing it cannot be done and doing it anyway, ideals before money — is exactly what others call "idealism." I reject the name but accept the description. I think and act exactly like an idealist would, but I will not use the word for myself.
This is not confusion. It is knowing that some words are too heavy. When others use them on you, they have already written the ending. And I do not accept that ending.
This page is not a manifesto. It is a record in progress.
The Journey
My name is Steve — because as a child I fell in love with Steve Jobs's iPhone 4. Technology is my native language. It is the intersection of science and art, the thing that gives me the purest instinctive joy.
But I was a humanities student. In Hong Kong, the best a humanities student could do was law. I studied it at City University — a good career, but not mine. Law is neither science nor art. It is a group of profit-driven people playing with words. I could not do well at something I did not love.
I went to HKUST for marketing. I love studying people — discovering what stays constant when everything else changes. But the workplace had no interest in that. They wanted sales, not research.
Here is an important distinction: marketing is not sales.
Marketing means studying users, understanding their real needs, and making products that truly serve them. Sales means closing deals for commission — you do not care whether the user needs it, only whether you can push it out the door. Most positions called "Marketing" in today's workplace are actually sales.
This distinction is my bottom-line judgment on every job: are you helping people, or are you taking from them? Get the stance right, and everything else follows.
Working for others has never made me happy. The ethos never aligned. Only when I build my own websites and software — using my own principles — do I feel alive.
First principle: structure before code. Get the skeleton right, and the flesh will follow. Website, strategy, market research — always structure first. If the structure is wrong, everything after it is patching. Second principle: technology serves the pursuit. Not the other way around.
Work Philosophy
These are not rules in priority order. They are different faces of the same stance:
The starting point is helping, not commission. The insurance agent is not hateful because they sell insurance. They are hateful because they pretend to help while actually taking from you.
What you say and what you do must match. People who talk one way and act another have achieved a self-serving coherence—but they harm others.
Xiaomi's logic is mine too. I can earn nothing, but I will not take a loss. And I do not work for free — free means you are the one paying.
The stance matters more than the outcome. Winning is best. Losing is not disgraceful.
Something that works well will naturally look good. If decoration communicates nothing, remove it.
Both is best. If you must pick one, pick the one that lets you sleep at night.
Tech & Ethics
I list a stack of technologies on my Tech page. But the stack is not the point. What you build with it is.
Automation is not harassment. If you ask me to scrape 10,000 personal records and send 100,000 spam emails — I cannot do it. Not technically. Ethically.
Recommendation algorithms cause harm. They keep people trapped in the internet, consuming time and attention. Technology should serve human growth, not corporate ad revenue.
Reliability is everything. Everything else can be debated, but what must not break must not break. Simplicity is aesthetic, but reliability is the bottom line.
Stack logic: TS for web, Python for scripts. Whatever fits the job. I do not care if the technology is the latest. I care if it fits.
If you build technology without caring what it does to people, you are not a real engineer.
Self · Others · Nature
When faced with a choice, ask first whether this is what you want to do, then whether it is useful. This is not selfishness. If I do not believe in what I build, who am I building it for?
I can spend three days perfecting an experience for a user I will never meet, but I regularly offend the person sitting in front of me with how I say things. Not on purpose. I focus on the principle, not on how you feel right now. This does not make me easy to get along with. But studying the market is studying people — it is the same ability looking in two different directions.
At Mu Xin's late-afternoon pavilion in Wuzhen, I understood something: humans are not nature's masters. Humans are inside nature. The brilliance of a Suzhou garden is not in how the rocks and ponds are arranged — it is that you never feel like you are "looking in from outside." You are inside it. I try to recreate this feeling in every design: a structure that feels like it was always meant to be there, not like you forced it into place.
No binary thinking — judge at the intersection
My training and influences are inherently dual. Chinese landscape painting gave me the instinct to subtract until only what matters remains. Bauhaus gave me form follows function — the same conclusion from a different road. Su Shi and Zhuangzi taught me abstract thinking. Unix philosophy taught me modularity and silence is gold.
This is not eclecticism. It is not "Eastern spirit, Western utility." Every judgment I make lives at the intersection — I use whichever tool fits, but the standard stays the same. That standard is not culture or tradition. It is: "what should this thing be, on its own terms?"
Not to keep everyone away. To save the wrong people a wasted trip.
This site is not for everyone. I said: "I do not want 10,000 visitors. I want the one who sees it and thinks 'this is right.'" The other side of that sentence: if you suspect you are in the wrong place, you probably are.
If you are any of the following, we do not work together:
Spammers — scrape 10,000 personal records and send 100,000 emails. I cannot do it. Not technically. Ethically.
Deceivers — the insurance agent who says "I am helping you" while taking half your commission. I can earn nothing, but I will not let what I do harm others.
Data thieves — other people's data is not your asset.
Also: I do not work for free. I can earn nothing, but I will not take a loss. Free means you lose — your time, your energy, your commitment to the thing. Someone who will not pay for your help does not really need it.
Principles shape products. See what these principles produce.
Explore my Blog →