I've been asked for advice by people a few years behind me, and I've noticed I give the standard answers — build projects, read code, be curious — which are true and have never once changed anyone's behaviour.

Here's what I'd actually say, if I were talking to myself at twenty.

1. Being stuck is normal, and staying quiet is the problem

I thought asking a question revealed that I couldn't do the job. So I'd burn a day on something a colleague would have unstuck in four minutes.

The thing I misjudged: nobody was tracking how many questions I asked. People absolutely did notice when something took a week.

The rule I'd give: thirty minutes of genuine effort, then ask. Say what you tried. That's not weakness, it's a status report with a question attached.

2. Read more than you write

My instinct was to start typing. Typing feels like progress.

Most of the expensive mistakes I've made would have been prevented by twenty minutes of reading first — usually discovering the thing already existed, or that the problem was different from what I assumed.

The best half-hour I've spent recently produced zero commits and cut a week of work to an afternoon. Nothing in any tracking system can see that.

3. Your fix is a hypothesis until the data agrees

This is the one I'd repeat loudest.

I used to build an explanation, find a change that matched it, see the symptom go away, and declare victory. Sometimes the symptom went away for an unrelated reason, and the actual bug shipped.

A theory that explains the symptom is not evidence. A local reproduction is not evidence. Evidence is the actual failing thing, with its actual logs, saying what you claim.

I still get this wrong. I got it wrong this year and it cost two days.

4. Nobody is coming to tell you what matters

I waited a long time to be told what to work on, and I was always given something, so the strategy "worked."

What it cost was judgment. Deciding what's worth doing is a skill that only develops by doing it, and if you never choose, you never build it. The people who seemed to be ahead weren't smarter — they'd been making calls, and being wrong, for years longer than me.

The one I'd skip

"Follow your passion." I'd replace it with: get good at something, notice that being good at a thing is enjoyable, and let interest follow competence. That's the order it actually happened in, for me.