Temple of Curiosity

Temple of Curiosity 002: the compounding issue

Temple of Curiosity — Mon, 12 Oct 2026 08:00:00 +0000

The weekend that paid off eighteen months later¶

a follow-up to this site's essay on why curiosity compounds.

A few readers asked for the source behind last week's linked essay. The short version: knowledge acquired for no immediate reason tends to arrive pre-connected to the question that produced it, which is exactly what makes it transferable later. Full essay here.

Why this matters – the takeaway: the right question for a curious detour is never "will this be useful," because you can't know. The right question is whether you're keeping a cheap, low-friction way to capture it when it does pay off.

Read more: Why curiosity compounds

Publishing cadence is a pacing problem, not a discipline problem¶

most "build in public" efforts fail from an unsustainable rhythm, not from a lack of willpower.

This week's site essay makes the case that a weekly-update promise is easy to make and brutal to keep once the interesting milestones thin out. The fix isn't more discipline, it's decoupling publishing cadence from the work's actual rhythm.

Why this matters – the takeaway: if your own update schedule has started producing filler, that's a signal to change the schedule, not to write harder.

Closing thought: issue 1 got more replies about the airline pricing item than anything else this quarter. Noted for issue 3.

Temple of Curiosity 001: three ideas worth chasing

Temple of Curiosity — Mon, 05 Oct 2026 08:00:00 +0000

Airlines have been running the same experiment on you for forty years¶

dynamic pricing isn't new, it's just gotten better instruments.

Revenue management, the discipline behind airline seat pricing, started in the 1980s as a defensive move against low-cost carriers undercutting fares. The surprising part isn't that prices move, it's how directly the underlying demand curve gets exposed once you know what to look for: the gap between a Tuesday fare and a Friday fare is a measurement, not a trick.

Why this matters – the takeaway: any market with perishable inventory and segmentable demand will eventually look like this. Concert tickets and hotel rooms already do. Worth watching for it anywhere "sold out" and "discounted" coexist in the same week.

Spaced repetition works, but not for the reason most apps advertise¶

the interval matters less than the act of nearly forgetting.

The learning research behind spaced repetition keeps finding the same thing: the benefit comes from retrieving a memory right at the edge of forgetting it, not from any particular schedule. Apps that space reviews by a fixed formula are approximating this, not guaranteeing it.

Why this matters – the takeaway: if a flashcard feels trivially easy every time you see it, the interval is too short and you're wasting reviews. The difficulty of recall, not the calendar, is the signal to watch.

A small essay on reading slowly¶

why the shape of an argument survives longer than its summary.

This issue's essay on the site argues that slow reading isn't about comprehension, it's about retention of structure over bullet points. Read it here.

Why this matters – the takeaway: if you only ever read in one sitting, you may be optimizing for having read something rather than for still having it.

Closing thought: every issue this year has had at least one item that started as an unrelated rabbit hole. That's not an accident; it's the whole premise of this newsletter.

Why curiosity compounds

Temple of Curiosity — Fri, 02 Oct 2026 09:00:00 +0000

Most advice about learning treats curiosity as a mood: something that shows up on good days and vanishes under deadline pressure. That framing misses the mechanism. Curiosity is not a mood, it's compounding interest on time spent exploring something with no immediate payoff.

The side quest that wasn't¶

A few years ago I spent a weekend reading about how airlines price seats, for no reason beyond finding one blog post interesting. It had nothing to do with my job. Eighteen months later, that same demand-curve intuition was the thing that let me see, in minutes, why a pricing experiment at work was doomed. Nobody could have scheduled that weekend as "training." It only paid off because I'd already done it before I needed it.

That's the pattern: the return on a curious detour is almost never visible at the time you take it. It shows up later, attached to a problem you haven't met yet, in a shape you wouldn't have guessed.

Why it compounds instead of just adding up¶

A fact learned in isolation is worth one use. A fact learned because you were curious tends to arrive already connected: to the question that made you look it up, to the other things you read around it, to the half-formed hunch it confirmed or killed. Those connections are what let it get re-used in contexts far from the one where you found it.

This is why curious generalists often out-predict narrow specialists on problems outside the specialist's lane: not because they know more, but because their knowledge is wired for transfer. Every new curious detour doesn't just add a fact, it adds more junctions the next detour can plug into.

Protecting the habit¶

None of this argues for chasing everything. It argues for treating unscheduled, non-instrumental reading as infrastructure rather than indulgence, and for noticing that the question "will this be useful" is the wrong one to ask about it in the moment. You can't know. That's the whole point.

Three things keep the habit alive in practice:

A standing, low-friction way to capture a tangent without derailing the task at hand, so curiosity doesn't have to compete with focus in real time.
Permission to read one layer below "need to know," even when nobody is asking you to.
Revisiting old notes occasionally, since compounding only shows up in hindsight.

Curiosity that compounds doesn't look productive while it's happening. It looks productive two years later, when the thing you learned for no reason turns out to be exactly the thing you needed.

The case for slow reading

Temple of Curiosity — Fri, 14 Aug 2026 09:00:00 +0000

Speed-reading culture treats books as a backlog to clear. The measure of a good reading habit becomes a number: books this year, pages this week. That metric rewards exactly the wrong behavior for anything you actually want to keep.

Two different jobs¶

Skimming and slow reading aren't the same activity at different speeds; they're different jobs. Skimming is for deciding whether something deserves your attention at all, or for extracting one fact you already know you need. Slow reading is for letting an argument rearrange how you think, which only happens if you give it room to disagree with you, stall, and get picked back up a day later with half of it still unresolved.

Treating the second job like the first is why so much "read" material leaves no trace. You cleared the backlog and kept nothing.

What slow actually buys you¶

The payoff of reading slowly isn't comprehension in the narrow sense, it's retention of the shape of an argument rather than its bullet points. A book read in one sitting compresses to a single impression. A book read over two weeks, with gaps where you argue with it internally, ends up stored as a structure you can reach for later in pieces, which is the only form that is actually useful when a related problem shows up.

Slow reading isn't a virtue in itself. It's a mechanism for converting attention into something durable, and it only pays off on material dense enough to reward the extra time.

Building in public without burning out

Temple of Curiosity — Fri, 22 May 2026 09:00:00 +0000

"Build in public" gets sold as a transparency practice. In most failures it's actually a pacing problem: people commit to a publishing rhythm sized for a launch week and try to hold it for a year.

The rhythm that breaks people¶

A weekly update is easy to promise and brutal to sustain once the interesting milestones run out and you're left reporting on plumbing. The honest fix isn't to fake momentum, it's to decouple the publishing cadence from the work's actual rhythm: some weeks produce something worth saying, most don't, and a schedule that pretends otherwise manufactures either burnout or filler.

What actually holds up¶

The practices that survive past the first few months share one trait: they cost little on the weeks when nothing happened. A running log kept privately, published in batches when there's a real shape to show, asks nothing of a slow week. A fixed weekly post with a word-count expectation asks everything of it.

Building in public is worth doing. It's just not worth doing on a clock set by someone else's launch month.