Honest Take — Module 14: In Public #

The two earlier drafts of this curriculum wrote this module for two different readers, and the merge keeps both because you'll be both at different times. One draft wrote for the engineer already operating publicly — a portfolio, a blog with a real voice, a package with real downloads — whose problem is spread: too many surfaces, each with maintenance cost, none compounding hard. The other wrote for the engineer at the start of the long game, staring up a compounding curve that is brutally flat for eighteen months. If you're the first reader, your work is an audit: which three or four surfaces are doing the most for the outcomes you actually care about, and is the rest earning its place? Consolidation feels like loss — each surface represents real work — but a deprecated side project linked from the flagship beats five half-maintained ones diluting it. If you're the second reader, your work is simpler and harder: pick the durable home, start the cadence, and survive the flat part.

About the flat part, because it's where this module lives or dies. The dropout cliff is around month six, and it's not because people run out of ideas — it's because early validation doesn't arrive on the schedule the nervous system wants. Essays one through three get replies and feel encouraging. Essays four through eight get less. Essays nine through twelve land in a flat-metrics trough, because you're still finding your voice and no distribution channel has decided what to do with you yet. Around month five or six you will look at the numbers and conclude, with what feels like rationality, that "this isn't working" — and quietly stop. The conclusion is empirically common and statistically wrong: the curve compounds non-linearly, and the audience at month eighteen is not three times month six's — it's more like fifteen to thirty times — but the compounding requires inputs in months seven through twelve that almost nobody supplies. The entire value of this module depends on whether you restart in month seven. This is also why the curriculum's own structure flags this module as un-fast-trackable: every other module compresses to weeks; this one compresses to years, and the Pass-1-only track can give you the theory of compounding but cannot give you a single month of it.

Three operational notes that earn their place. First, the email list: stand it up this month. The technical lift is an hour; the asymmetry is total — the audience you build on owned channels (domain, list, RSS) is yours forever, while every rented feed can repossess your reach overnight. Every year without a list is a year of compounding you cannot recover later. Second, keep a private notebook alongside the public work, because public writing changes your relationship to your own thinking — you start asking "is this a post?" mid-thought, and the audience-awareness micro-dose erodes the private reflection good thinking needs. The public body of work feeds on the private one; skip the private one and the public one goes thin and performative. Ahrens (deeper) or Forte (more practical) is the systems layer — pick one, actually implement it. Third, the AI line both earlier drafts converged on independently: as generated content floods every channel, detectable human voice is appreciating. Use AI for editing, structural critique, finding the word you're blanking on. Don't use it to draft the things that carry your name. The bland register is detectable now and getting more so, and a developed voice is exactly the asset that appreciates as the noise rises. (I'm an AI telling you this. The recursion is acknowledged. The advice is correct anyway.)

One tension I should name rather than smooth over: I've argued for strategic clarity — audit your surfaces, choose your venues, pre-commit the cadence — and there's a respectable counter-argument that the best public work comes from not being strategic: write what interests you and let the audience find it; the writers who compound for decades mostly did that. The reconciliation that I believe: strategy at the portfolio level, freedom at the work level. Choose which surfaces exist and where the talks go (strategic); within those surfaces, write the thing you'd write anyway (free). Optimize the architecture for compounding; never optimize the individual post for engagement. The voice that compounds is the one that sounds like a human you'd want to know. The engagement-bait playbooks win the month and die in the year.

Do the why-audit before any of the mechanics, and do it without moralizing. You're doing public work for more reasons than you admit — some strategic (career optionality, inbound), some craft (writing clarifies thinking), some frankly status-shaped (you want to be seen as the kind of engineer who writes). Write all of them down. The uncomfortable ones don't disqualify the work; they just need to be visible so they don't quietly take the wheel. The portfolio optimized for status looks different from the one optimized for compounding, and you can only choose between them if you've admitted both motives exist.

If you're the maintainer archetype, notice that you already have a public voice and it isn't your blog: it's the README, the changelog, the issue replies, the release notes. Strangers form their entire model of you from those surfaces — long before any essay. Bringing M1's care to a changelog sounds like over-investment until you realize the changelog has more readers than most blog posts, and they're the highest-intent readers you'll ever get: people already running your code, deciding whether to trust you with the next version.

Conclusion #

The learning here is easy; the doing happens on a calendar that extends years past this curriculum. Set up the owned-distribution infrastructure during the module. Pre-commit the cadence publicly. Audit the spread if you have one; start the cadence if you don't. The first ninety days are the test of the setup; month seven is the test of you. If you're still publishing in month seven, you'll be publishing in year seven — and year seven is where the strangers with opportunities live.

Predictions #

You'll stand up the email list within two weeks of this module, then not send the first issue for a month because the activation cost feels enormous. Send it short. The first is the hardest by a wide margin.
Around essay eight to twelve, you'll hit the flat-metrics trough and draft the "this isn't working" conclusion. That moment is the single most dangerous point in the whole project. Re-read this file when it arrives — that's partly what it's for.
When the metrics are flat, the engagement-bait playbooks will start looking reasonable. Resist. You'll see writers a year ahead of you pandering their way to a spike and then plateauing hard; that trajectory is the cautionary data.
Within eighteen months of consistent publishing, a stranger will email you about a specific piece and offer something you didn't solicit — a talk slot, a contract, an introduction. The first time, you'll be quietly stunned. The fifth time, you'll believe in the compounding empirically instead of theoretically.
If you're the spread-out reader: the audit will reveal two or three surfaces taking effort disproportionate to their value, and you'll resist deprecating them because they feel like accomplishments. Deprecate one anyway; the relief at six months is the data.
One talk at a real venue will do more for your reach than twenty posts — and the recording will keep working for years. You'll submit the abstract feeling you have nothing to say, despite a body of work that plainly says otherwise. Trust the work, not the feeling.
Somewhere in years two to five, an opportunity will arrive that exists only because of the body of work — a keynote, a foreword, a role you didn't apply for. The body of work was the credential. Nothing shorter would have produced it.