career

Frequent change hurts most when context never has time to become something you can carry into the next environment.

Instead of predicting one hot job years ahead, combine foundations, current tools, real contexts, and multiple exits.

One month's revenue is only the start; duration, concentration, hours, and net cash matter more.

In 2022, US job switchers out-earned stayers by 2.6 points; in the first half of 2025 it flipped. The real cost of staying put is not knowing what you're worth right now.

A favor that costs nothing on paper isn't free — the price just doesn't show up on the money ledger. It shows up on the other ledger, the one for trust and default expectations, and nobody ever tells you that one's running.

I fixed two bugs in my own writing-automation system this week — an image retriever that returned zero images, and a dedup logic that couldn't tell facts from opinions apart. The real question was never whether it was worth it. It was whether I was willing to pay for it.

An AI side-hustle course dropped its price from 13,400 yuan to 980 in a few days, with one class manager sending 36 nudge messages in 36 hours. That's a psychological game, not a discount — and it points to a better question than "does the method work?"

Whether someone actually helps you has nothing to do with how close you are — it's about whether the favor touches their own interest. They know this math better than you do.

New Stanford data: in AI-exposed jobs, employment for 22-25 year-olds runs 19% below their peers, and the gap keeps widening every month. It's not that your skills got worse — the rung you used to practice on is being pulled out from under you.

Loyalty resets to zero the moment you switch companies. What actually travels with you is the judgment that lets you see a problem before everyone else does — backed by a look at a 1959 theory of the five bases of power.

Background isn't only money and rank — insight, trust, and judgment are a kind of background too. You might already be holding it and not recognize it.

After scrolling through my own raise history for the last few years, I finally sat down and looked at the cost sheet every boss stares at: headcount is one row, compute is another, and they point in opposite directions. One drops tenfold a year, the other creeps up a point. Why grinding your skills harder doesn't get you off that sheet, and what you can actually change.

China has about 200 million people in flexible employment, 43% of the urban workforce. 'Flexible' describes the company's staffing flexibility, not your freedom at work; the head start worth grabbing is delivering something on your own, early.

An exit line is a trader's term: before you get in, decide where the loss makes you walk. A job needs one too. It is easy to mistake the operation growing for yourself growing, and loss aversion plus "just a little longer" keeps you renewing year after year. Do the books once a year: did salary, deliverable skill, and your way out actually move.

A delivery pipeline of mine was marked 'verified' for two days without ever touching the line of code that could fail — until it finally crashed on a real batch of 27 records. It made me realize: as teams shrink and AI takes over more of the work, the free layer of humans catching your mistakes disappears with them.

The job hadn't changed in three years — what changed was who I measured it against. Mixing those two up is how people talk themselves into an irreversible decision for the wrong reason.

Worth isn't measured by how much you can do alone, but by how much you can pull together into a plan that actually works. A daily writing pipeline taken over by AI agents made me rethink how I'd been scoring myself.

I fixed a database outage at 2am. The next morning it got summed up in one line: "the team already handled it." Doing the work well doesn't matter much if nobody says what you actually did.

The people who get ahead at work aren't always more skilled — they've just studied the players in the room. The difference between fundamentals-only and reading-the-game isn't effort, it's one extra question.

Whether a side hustle is worth your time isn't about how busy it keeps you. It's about whether finishing it once means you have to start over next time — the difference between a one-off script and a service that keeps running.

If you only finish your own slice, you never see how the game is actually played.

Being accommodating is not a morality question. It is a ledger: good temper that buys nothing back is money quietly leaving your account.

To judge whether a side project or freelance job is solid, do not look at how busy it keeps you. Look at whether you can explain how you plan to do it, why it should be you, and what stops others copying it once it works.

Flash developers vanished five years ago, not because the technology was bad, but because the platforms turned together. This time it is AI's turn.

Ten years as a tech manager, and going through the ledger shows most of my experience is tied to one company's walls — it zeroes out when the badge goes back. What you take with you is the part you can deliver to a stranger without first proving where you worked. AI amplifies this question rather than letting you avoid it.

An electronics factory in Ningbo fines workers 3,000 yuan for leaving between July and October. Lawyers say it is likely illegal. The reason it can be written down at all is not that it holds up, but that whoever wrote it calculated you probably won't walk.