A Demand — The Part I Can't Say Yet
There's a demand right now that I can't describe in any real detail. It's high enough that meeting it means "scaling." Not a metaphor, a literal backend problem, the unglamorous kind that doesn't photograph well and can't be threaded. I've sat with it for weeks now, and the honest answer to "how's it going" has mostly been: it's coming. Not because I'm being coy. Because there's nothing sayable yet that wouldn't be premature.
This blog has quietly gone through its own version of the same shift. In the first months I drafted ahead, building a runway of posts before I needed them, the way you'd stockpile before a storm. Lately I don't. Not because the well's run dry, but because writing here has folded into everything else I'm carrying, and I've stopped needing the pipeline to prove the commitment is real.
A Restraint — Sitting On It
A few weeks ago, an announcement went out. A partnership with a global institutional fund, the kind of news that, in this space, usually gets front-run by days of "big things coming" posts before it's even official. I felt that pull. The news was real, it was mine to know early, and there's a version of me that would have posted around it: hinted, teased, let the room fill in the blanks.
I didn't. Not out of restraint for its own sake. Out of a distinction I keep having to relearn: urgency and haste are not the same thing. Urgency is knowing something matters. Haste is needing everyone else to know it matters right now, on your timeline, before the thing is actually ready to be known.
A Split — Not the Same Muscle
Here's the part that surprised me. AI hasn't made me want to move faster. It's made me want to build more: more drafts, more iterations, more of the unglamorous scaffolding work that used to eat entire weekends. That part is genuinely liberating. Do more with less; I believe that now in a way I didn't a year ago.
But publicising it is a separate muscle entirely, and the tool has nothing to say about that one. It amplifies whatever you're already pointing at: output, iteration, speed of build. It has no opinion on whether you should tell anyone. That decision is still fully, uncomfortably mine, and the temptation to publicise doesn't shrink just because the work moves faster. If anything, there's more to show, more often, and the discipline has to work harder to keep up.
A Discipline — The Long Correction
I remember an analogy from my late teens, early twenties: discipline as doing the thing you don't want to do. Simple, blunt, useful for a while. Somewhere in my thirties it turned into something quieter: self-discipline, which isn't about forcing the unwanted thing anymore, it's about not needing an audience to do the right thing at all. That correction is still underway. I won't pretend it's finished. But I'm doing better with it than I was six or seven years ago, and I'll take that.
Most of what this actually looks like, day to day, is repetition that feels like standing still. The same review, the same follow-up, the same unglamorous pass at something nobody will ever see finished. It's grinding, not in the way anyone puts on a highlight reel. And I've stopped assuming the people who build loudly are doing it for the wrong reasons. I don't actually know their intent, and pretending I do is its own kind of dishonesty. What I can control is mine: head down, stay humble, let the work do whatever talking it's going to do.
A Distinction — What Can't Be Taught
Everyone can be taught to build. Given enough time, almost everyone can build. But not everyone can actually build: the kind that holds up when no one's watching it happen, the kind that doesn't need the room to know it's happening at all.
That difference isn't a skill. I don't think it's teachable. I think it's chosen, quietly, most days, long before anyone sees the result.
IN.
Previously → The Architecture of Trust