This is the one with the cleanest timeline. Eleven actual weeks, start to finish. I had a landing page, a payment processor, six testimonials I’d extracted from clients with the gentle persistence of a timeshare closer, and a little animated checkmark that pulsed every time someone hovered over the buy button. I was unreasonably proud of that checkmark. I’d built every component of a functioning business except one: a human being willing to trade money for it.

(The checkmark, to its credit, performed flawlessly. It pulsed at no one, all day, every day, like a lighthouse somebody’s still paying the electric bill on, for a coastline with no ships and no plans of getting any.)

What had I done wrong? I guessed. I sat alone in a room and decided what my audience needed with the rigor of shaking a Magic 8 Ball until it landed on the answer I already wanted. Earnestly. Confidently. With zero supporting evidence.

Here’s the part that still bothers me several months later, the way you think of the perfect comeback to an argument long after the argument’s over: my audience had been telling me exactly what they wanted the whole time.

Nobody Asked You to Build That

My eleven-week mistake wasn’t a one-time lapse in judgment made only by myself. It’s the default setting for how most people decide what to build next.

You get a hunch. The hunch feels solid, mostly because it’s yours and you’ve been marinating in it for weeks, which is a terrible test for whether an idea is good and a great test for whether you’re attached to it. You build the thing alone, behind a closed door, the way you’d defend a thesis nobody’s allowed to read until the day it’s due. Then you publish the offer.

And what comes back isn’t pushback. It isn’t even a polite no.

It’s crickets. You’d have gotten a better response standing at the edge of your driveway with a poster board, yelling random prices at oncoming traffic.

And here’s the part that should make you want to lie face down on the kitchen floor for a minute: the answer was sitting in your replies the entire time. You walked right past it on your way to go build something nobody asked for.

Substack Notes are not just a place to post half-baked ideas and refresh for the dopamine hits. (Several thousand accounts would loudly disagree with me on this, several times a day, but several thousand accounts are also the problem.) Used correctly, Notes are a live, ongoing, completely free research panel made up of people who’ve already opted in to caring what you think. The signal is there. Almost nobody is reading it as signal. They’re reading it as a content calendar.

Here’s the actual difference, in one sentence: a Note that performs well tells you what’s interesting. A Note that gets real replies, the “I’m dealing with this exact thing right now” kind, tells you what to build. Most people only ever go looking for the first one.

So before you build anything else, here’s how to actually use the thing you’re already sitting on.

The Cure for Crickets

Here’s the cure, in six steps. It starts with choosing one problem with actual money attached to it.

Step 1: Pick a problem with a pulse. Not "manifestation." Plenty of people buy "manifestation." Somehow the universe keeps not delivering, and somehow that's never the methodology's fault. You want a problem with a pulse and a price tag. Something like: I've lost the same fifteen pounds three times and gained it back every single time. That sentence has rent attached to it. “Manifestation” has a vision board attached to it, and I kept mine up for two years. The rent situation never improved.

Step 2: Publish ten angles on it. Take that single problem and poke it from ten different directions, watching closely to see which poke makes somebody yelp.

Diagnose it. Take the unpopular position on it. Confess the dumb mistake everyone makes with it, yourself included, because that one always lands. Tell the tiny story of the time it bit someone in the tookus. Ask the blunt question nobody else is asking. Walk through the boring process. Say the objection out loud that everybody’s too polite to admit. Name the consequence of ignoring it. Lay the before and the after side by side. Dare somebody to tell you you’re wrong.

That’s ten, and no, you’re not fishing for the cleverest sentence. The cleverest sentence is a trap. It earns applause and teaches you nothing. You’re fishing for the framing that makes a stranger stop scrolling, and wonder, seriously, if Google sold your personal search history again.

Step 3: Write down the language that comes back. This is the step almost everyone skips, because it requires shutting up and paying attention.

Which Notes pull actual subscribers, not just claps. Which question keeps bobbing to the surface of your replies like a cork you can’t hold under. Which comment opens with “I’m dealing with this exact thing right now” (mark that one, it’s a flare). Which objection refuses to die no matter how many times you bury it. And, most valuable of all, which phrases your readers use that you’d never put on a landing page yourself, because you’ve been marinating in your own industry’s vocabulary so long you’ve forgotten what an actual human sounds like. (This isn’t unique to you. A debt counselor says “debt consolidation strategy.” Their actual client says “I don’t know which card to pay off first and I’m scared to look.” Yours is happening too, you’ve just stopped being able to hear it.)

That last one is the entire prize. Product research you didn’t have to commission, gathered from the only focus group that’s ever told the truth: the one that doesn’t know it’s a focus group.

(A word on surveys: don’t, or at least not first. A survey is where honest answers go to quietly expire. The instant a question lands in a form, people stop telling you what they want and start telling you what they suspect a reasonable person ought to want, which is a different thing entirely and useful to nobody, least of all the reasonable person, who as far as I can tell has never bought anything from anyone.)

Step 4: Promote the winner into long-form. Once a framing draws blood, you build on it. The Note proved the nerve exists. The article does the actual work of digging that nerve out and turning it over under good light. The comments on the article then tell you, with almost rude precision, what you still haven’t answered.

Step 5: Build the smallest sellable thing. A checklist. An audit. A template. A live teardown. A one-page diagnostic. Not a forty-three-module course filmed alone in a closet at two in the morning with the production values of a hostage video. Something narrow enough to ship by the weekend and useful enough that a stranger pays for it without being asked twice.

Step 6: Watch the room sort itself. Here’s what actually happens once a problem’s been named clearly enough: people stop being an audience and start being applicants.

The people who reply, subscribe, argue, buy, and come back with the follow-up question are sorting themselves. By intent. In public. Directly in front of you. You didn’t assemble a cold list and pray over it. You didn’t buy a thousand email addresses and interrupt a thousand dinners. The interested ones walked toward the thing that named their problem and stood there, waving. You stop guessing who your market is, because you’re watching it assemble itself, one self-selected human at a time.

I still think about that checkmark sometimes. Pulsing away, doing its job perfectly, for nobody. The product never sold. I shut it down eventually, but the checkmark’s logic didn’t, it’s just aimed correctly now. The next thing I built started with a question instead of a guess, and I asked it out loud, in public, before I wrote a single line of anything.

Ask first. Build second. Everything else on this list is just logistics.

Crafted with love (and AI),

Nick “Unpaid Focus Group of One” Quick

PS... If your AI keeps sanding every Note down into the same oatmeal-toned LinkedIn voice, that's not a content problem. That's a voice problem, and it's the one thing all this audience research can't fix for you on its own:

Get The Voice Sync System (FREE)

The system walks you through getting the machine to sound like you instead of every other account currently typing “consistency compounds” into a sunset. Or keep building on hunches and hoping the next one lands. The crickets aren’t going anywhere either.

PPS… If you want to test this without building a single thing, hit reply or leave a comment and tell me the one problem you’d run through this system first. I read every reply, partly to help and partly because, yes, you just became part of the research panel whether you signed up for it or not. (The transparency’s part of the brand. Allegedly.)

Here’s what it actually did.

Every one of those three legs has two halves. A cheap one and an expensive one. For years you could lean on the cheap half of each and mostly get away with it. Then the machine showed up, looked at the cheap half, and said “oh, I can do that,” the way it says it about everything, with the serene confidence of a guy who has never been in a fight explaining, in detail, how he’d win one, having just, last week, seen his first Steven Seagal movie.

So the framework still stands. The ground under it turned into a chocolate vanilla swirl pudding pack.

Let me introduce you to Brennan.

Brennan posts six times a day. He has a streak counter he’s mentioned in eleven of those posts, most recently to note that the streak is “a non-negotiable.” His bio says “AI-powered storyteller,” which is doing a lot of work for a guy whose last three posts were a Monday motivation, a repackaged thread about compounding, and a photo of his desk with the caption “environment is everything” (the desk has a Ring light, a $40 succulent, and a book he has not opened facing outward for the cover). His pinned post is about his morning routine, which involves single-origin Slovenian spring water served at precisely 13 degrees in a copper cup he refers to exclusively as “the vessel,” a $68 leather journal he photographs more than he writes in, and a sentence attributed to Marcus Aurelius, which Brennan read on a LinkedIn carousel eight months ago and has since come to believe he thought of himself.

Brennan is doing everything the advice told him to do. He’s also about to get quietly taken apart by all three legs of his own favorite framework, because Brennan mastered the cheap half of each one, and the machines just made the cheap half free.

Here we go. One at a time.

Know

“Know” used to be the leg that hurt. Getting found was the whole bottleneck, so it earned a permanent seat at the strategy table and everybody treated it with the reverence of an unwritten rule traced back to a Vine nobody can find from an account nobody can remember since the app left the Play store in 2017.

Now a faceless account churning out AI grannies reacting to 90s hip-hop gets known. Bots get known. (Awareness used to be a moat. Now it’s a bouncer who stopped checking wristbands.)

Every modern version of this advice quietly walked “know” out of the strategy room, patted it on the head, and reassigned it to logistics. Post more. Hit your quadrants. Six videos a day, one per topic, repeat until something sticks or you do.

Brennan is winning this leg, by the way. Brennan is extremely known. If you mute him he finds you in a different app.

There are two things hiding in “know.” Getting known means a stranger registers you exist. Getting recognized means a reader clicks on name alone. AI buried the first one. It can’t touch the second.

Like

“Like” is where it gets interesting, because this is the leg most people chop off themselves, proudly, with a hacksaw, a LinkedIn post about why they did it, and a follow-up thread three days later about the freedom they felt afterward.

Your gym selfies were never a business. Everyone gets that part. (Brennan’s gym selfies are also not a business, but there are nine of them this week, each captioned with a Stoic insight Brennan is fairly certain he arrived at independently, having no memory of the podcast it came from or the macro-planning session it soundtracked.)

But “like” was never only the selfie. There are two versions hiding under one word. One is relatability theater. The other is sensibility: the specific texture of how one particular brain runs at the world. The second one is the entire reason an idea gets repeated in a room you will never be let into.

And here’s the part that should land harder than it does. The machine is exceptional at theater. It’ll hand you a relatable morning-routine post, a tender little origin story, a “real talk, I almost walked away from all this last year” confession, on demand, in bulk, faster than you can feel an actual feeling about your own actual life. The relatable half of “like” is free now. What it can’t fake is the sensibility, because sensibility isn’t a content type. It’s a person. You can’t prompt your way to that any more than you can prompt your way to a childhood.

Trust

“Trust” is the leg everyone’s amplifying right now, and almost everyone is amplifying the version that’s quietly being discontinued.

The pitch: become the expert, know the most things, trust follows. Solid plan in 2019. The clankers know more things than you now. Instantly, drafting up think pieces with footnotes, at three in the damn morning, while you’re asleep dreaming of puppy dogs and rainbows. Expertise-trust is finished.

Brennan has Top Voice energy and a slide deck and a way of saying “let’s unpack that” that makes you want to repack it and GTFO. None of it survives ten seconds against a chatbot that’ll unpack the identical thing for free and won’t make you sit through the part where he relates it back to the three weeks he spent living abroad at an all-inclusive resort in Cancun.

The trust that survives isn’t expertise. The machine has expertise. It isn’t relatability either. The machine has that on tap. What it doesn’t have is scar tissue. Decisions made under real conditions, with something actually on the line. Not what you know. What you’ve bled to figure out.

Now stand back and look at all three legs together. The same thing happened to each one.

Two halves apiece. Known and recognized. Relatable and singular. Credentialed and scarred. For years the cheap half of each was a workable shortcut, and most people, Brennan very much included, took it, because it was right there and it didn’t require the inconvenience of having a self.

Then the machine showed up and did the cheap half for free. It gets you known. It does relatable on command. It out-credentials you before breakfast, with footnotes, (again with the goddamn footnotes!) Every easy path to every leg, automated, handed to anyone with a pulse and a password, worth precisely nothing now for exactly that reason: everyone has access.

What’s left standing in all three is the same thing. The judgment call nobody else would have made the same way. The machine can’t supply it because the machine has no stakes. It has never once been wrong in a way that cost it anything.

A line. A frame. A way of seeing the thing they can’t un-see now.

If they carry nothing, you wrote the cheap half, the one the machine hands out at the door like a flyer for a gym that already closed. If they carry one sentence and repeat it to a friend three weeks later and forget it came from you, that’s the expensive half doing its job. That’s recognition. That’s the part no model ships, and the part Brennan, six posts a day and climbing, has never once produced by accident or otherwise.

Know, like, and trust still work. AI just automated the cheap way to earn all three, which leaves exactly one currency the machine can’t print, can’t fake, and can’t borrow from you: your taste.

Crafted with love (and AI).

Nick “The Expensive Half” Quick

PS… I publish here regularly. If this landed, subscribe, restack the line that hit, and forward it to the Brennan in your life who needs to read it. Then:

Grab the Voice Sync Framework free

Because the expensive half doesn’t survive on instinct. It survives because you documented it before AI had a chance to sand it down to something everyone would approve of. Brennan doesn’t have one. That’s not a coincidence.

PPS… Brennan read this far. Brennan is nodding. Brennan is already opening a fresh post titled “Why Taste Is the Real Moat in the Age of AI,” which he will publish in forty minutes, schedule across four platforms, and never once suspect was about him.

Great advice for checkout flows. Genuinely catastrophic for publishing.

There’s a whole optimization playbook built around friction removal. Short paragraphs. Scannable bullets. The pre-chewed “key takeaway” callout that deposits the insight directly into the prefrontal cortex without asking it to do anything. Carousels that sequence your argument so you can receive it while thinking about something else entirely. (The optimization industry calls this “reducing cognitive load.” I call it ambient content.)

Scrolled. Nodded at. Filed under “vaguely interesting, can’t recall it later this evening.” The reader is unchanged.

Slop factories run on zero friction. They need content that’s easy to consume because their model requires volume, and volume requires interchangeability. Anything that makes a reader pause, actually wrestle with an idea, or exert their mental faculties for three seconds is overhead in their system.

For us, that pause is the product.

Attention Doesn’t Compound

There’s a spectrum to what publishing can actually do to a reader.

At one end: they stopped. Saw the thing. Maybe nodded. That’s attention. Measurable, farmable, worth approximately as much as a door being open. Something might walk through it. Usually doesn’t.

At the other end: the idea travels. Your reader explains your concept to someone who’s never heard of you, in their own words, in a conversation you weren’t part of. That’s not attention. That’s what attention was supposed to be a door to.

You can’t optimize your way there.

Ideas only travel when readers carry them. And readers only carry ideas they earned. Ideas they had to piece together. Ideas that felt, in some partial way, like their own discovery. (There’s actual research on this, which is satisfying because it means I’m not just being contrarian for aesthetic reasons. I’m being contrarian for aesthetic reasons that also happen to be correct.)

Handed insight is forgotten insight. Earned insight gets carried.

That’s the mechanism. And it points directly at a strategy most creators are too worried about open rates to run: making your content slightly harder to consume.

Not confusing. Not opaque. Just requiring a little assembly.

Assembly Required

Research on the IKEA Effect shows people consistently value things they assembled themselves more than identical pre-assembled versions. They invested effort. The thing is partly theirs.

Content works identically.

When you give readers the principles and context behind a complex idea, then leave the final synthesis for them to close, something shifts. They feel the “a-ha” as their own discovery. They remember it differently. They explain it to other people in their own words, not yours. In a conversation you weren’t part of, to someone who’s never heard of you.

Samin Nosrat did this with an entire book. Salt Fat Acid Heat doesn’t give you recipes. It gives you the four principles that generate any recipe. Once you understand what salt enhances, what fat carries, what acid brightens, what heat transforms, you don’t need her instructions anymore. You can figure it out yourself. (That’s the point. That’s always been the point. The book sold millions of copies because it made readers feel like cooks, not like people following directions.)

She gave up the recipe and kept the kitchen. Most creators won’t do that. Giving readers the principles means accepting they might not need the next post. (Turns out readers who don’t need your instructions come back for your judgment. Which is the better thing to be selling anyway.)

Four Things to Leave Out

Not all friction is intentional. Some of it is just bad writing.

Four places to stop short. The reader takes it from there.

The Gap. Lay out the context, the principles, the evidence. Then trust the reader to make the leap. Don’t explain every implication. Creators who walk you through every consequence of every idea are treating you like you need protecting from your own cognition. Readers feel that condescension. They just don’t often name it out loud.

The Breadcrumb. Reference something significant without the full context. Call it “The Tuesday Anomaly” and let the curious go looking. You’ve just created a binge loop, a reason to read backward through the archive, and a quiet act of self-selection. The readers who go looking are the ones who care. (Those are your actual readers. The ones who don’t go looking are also fine. They just aren’t who this is for.)

The Invite Without the Answer. Leave a real question hanging. Not rhetorical. Not performative. “I don’t know how to resolve this yet. What would you do?” isn’t weakness. It’s a gap the reader steps into. Comments become generative instead of decorative.

The Earned Access. Put your most useful resource one deliberate step away from obvious. Not a treasure hunt. One intentional action. When they find it, they value it differently. The effort is the signal. It also quietly filters out the people who won’t use it anyway, which is its own kind of service.

The Filter Is the Feature

Friction isn’t just an engagement strategy. It’s an audience quality filter.

Optimize for zero friction and you’re optimizing for anyone. Which sounds like growth. Until you realize “anyone” includes every reader who’ll never buy, never refer, never reply, never come back unless the algorithm taps them on the shoulder and says “hey, remember this guy?”

The slop factories need everyone. They’re running volume.

You need the readers who'll take your framework to a party you weren't invited to. The ones who absorb something you wrote so completely it dissolves into how they think. No citation required. The idea doesn't travel as a quote. It travels as a new instinct.

Friction self-selects for those readers. Not through difficulty for its own sake. Through the implicit signal that this publication respects your intelligence enough to leave some of the work to you.

Slop factories have perfected the frictionless experience. You can’t beat them there. You shouldn’t try. Their model runs on volume, and volume can’t afford to slow anyone down. You can.

Crafted with love (and AI),

Nick “Licensed Friction Inspector” Quick

PS... You used to write like you. Now you write like ChatGPT writing like you. Engineered friction only works if there’s something underneath worth earning. Voice Sync documents what that is before AI finishes averaging it out. Four prompts. Twenty minutes. Free:

Get The Voice Sync Guide

PPS... If this earned it: like, restack, forward it to the creator who’s been optimizing themselves into ambient content. If it didn’t, the comments exist for a reason.

Not the weakest paragraph. The last one.

Writers add it for themselves. It’s the exhale after the work is done. A quiet summary of what just happened. A soft landing. Sex followed by a debrief. “So what that was, was intimacy. We connected emotionally and physically, which research suggests is important for long-term bonding.” (Nobody wants this. Nobody has ever wanted this.)

AI Does It Too. And It’s So Much Worse.

AI does it enthusiastically. Pathologically, even.

It was trained to. Human raters marked responses as “complete” when they had a tidy wrap-up, so models learned that a summarizing conclusion equals a finished piece. Now every AI output ends with some variation of “Ultimately, X matters because Y” or “The takeaway here is...” or the always-popular nothing-sentence that restates the entire piece in case you forgot what you just read.

The phrasing rotates. The move never changes.

Which means if you’re co-writing with AI and not cutting the last paragraph, you’re not just over-explaining. You’re leaving the most recognizable AI fingerprint in the entire piece sitting right where everyone looks last.

Your Reader Already Got There. You Just Don’t Trust Them.

The piece already made its point. The last paragraph just confirms it. Which tells the reader you didn’t trust them to get there on their own.

They noticed. They always notice.

The 30-Second Test That Proves It

Delete your last paragraph. Read the piece again. If it ends stronger, leave it deleted. If something’s genuinely missing, put it back.

Most of the time? It ends stronger.

Because the real ending was almost always the second-to-last paragraph. The last one is you sticking around after the party’s over. Or AI doing what it was rewarded to do. Either way, same problem. Same fix.

The Bow Is for You. The Summary Is for the Model. Neither Is for the Reader.

Writers over-explain because the work felt incomplete without a bow on it. AI over-explains because a committee of human raters told it “complete” looks like a summary. Both are writing for themselves. The reader finished the thought. They’re already standing up to leave.

Let them go.

Crafted with love (and AI),

Nick “Leave Before the Bow” Quick

PS... If this post made you delete your last paragraph, you’re already doing the thing. The next thing: making sure what AI hands back doesn’t quietly sand down the paragraphs you kept. Voice Sync is four prompts, twenty minutes, and it builds a Voiceprint that catches AI smoothing you down in real time. Free. File lands in 30 seconds. The only cost is finding out I was right:

cowritewithai.com/voice-sync

PPS... If this was useful: subscribe, like, restack, the full set. Then open something you wrote this week and delete the last paragraph right now, before you talk yourself out of it. Come back and tell me what happened. That’s the comment I actually want.

You imagine you’ve hired a tiny, stupid goblin to write the first version of whatever you’re working on. He’s underqualified. His dialogue is terrible. He appears to have learned punctuation from several unrelated ransom notes.

None of this matters.

You paid him doodley-squat.

The goblin’s job isn’t to write something good. His job is to get to the end, preferably without eating the keyboard. Then you come back, survey the damage, and turn whatever he left behind into something you’d admit was yours.

It’s ridiculous.

It’s also a very good cure for the blank page.

Not because bad writing is secretly good.

Because bad writing gives you something the blank page refuses to provide.

Something to push against.

The Blank Page Asks Too Damn Much

A blank page asks an unlimited question:

What should this become?

That sounds innocent enough. It’s not.

Now you have to choose the opening, angle, structure, tone, examples, level of detail, and exact point before a single sentence has earned the right to exist. You’re inventing and evaluating at the same time, which is roughly like learning to drive with someone grading your parallel parking before the engine starts.

No wonder the cursor stops.

You aren’t merely trying to write a paragraph. You’re trying to write the paragraph your readers will see. The one that proves you still know what you’re doing. The one that should be original, useful, funny in precisely the correct dosage, and somehow arrive fully dressed before you’ve worked out where the piece is going.

A perfectly ordinary opening sentence arrives dragging a parade float shaped like your entire reputation.

Then we act surprised when it can’t move.

The goblin has no such problem. He has no subscribers. No respected peers. No internal committee asking whether sentence two is sufficiently original to justify its continued existence.

He just starts typing.

The Goblin Changes The Question

Once something exists, the problem gets smaller.

You’re no longer asking:

What should this become?

Now you can ask:

Is this the right angle?

Is the opening too broad?

What’s missing?

Why do I hate option four?

What does option seven almost understand?

Those are easier questions because writers are often better at reacting than originating. Give me a blank page and I can stare at it until one of us develops a personality disorder. Give me ten questionable openings and I’ll know within minutes that six are wrong, two are boring, one contains a useful question, and the last one has accidentally tied the whole piece together.

That’s what Goblin Mode actually does.

It turns writer’s block from a generation problem into an editing problem.

AI happens to be very good at making that conversion quickly.

The Modern Goblin Is Annoyingly Competent

The metaphor breaks a little when you hand the job to current frontier models.

AI isn’t stupid. It can produce clean, coherent writing in seconds, often better than the average person could produce cold. The average person doesn’t regularly write for publication. AI has absorbed more competent prose than any human could read before the sun expands and newsletters become a secondary concern.

So the modern goblin is articulate.

Fast.

Educated.

And completely free from the emotional baggage you’ve attached to beginning.

It doesn’t need the opening to prove anything about itself. It won’t stare at sentence one for forty minutes because sentence one may determine how strangers perceive its entire body of work. It doesn’t open the refrigerator six times while hoping an angle materializes behind the Dijon mustard.

It can simply make something exist.

Sometimes the result will be weak. Sometimes it’ll be irritatingly competent. Occasionally it’ll contain a sentence, structure, or turn you couldn’t reach while supervising every word like a mall cop guarding the coin-operated kiddie helicopter.

All three outcomes can be useful.

The goblin’s job isn’t to lower the quality of the writing.

It’s to lower the cost of finding a direction.

Give It The Embarrassing Job

Take one real idea you’ve been avoiding.

Not an empty topic you hope AI will somehow make meaningful. A worthwhile idea you already have, but can’t find your way into.

Give the model the topic, the intended reader, and the rough point you think you’re making. Then ask for ten genuinely different approaches. Different structures. Different angles. Different openings into the same room.

Add one restriction:

Generate ten genuinely different approaches to this idea. Vary the structure, stance, and angle. Don’t rank them, combine them, recommend one, or imitate my voice.

Then read what comes back.

Don’t search for a finished post.

Search for a reaction.

Too broad.

Wrong story.

Good question.

Keep that line.

That response is the value. The model has turned one enormous creative demand into a series of specific decisions you already know how to make.

Choose the useful move yourself. Rewrite it. Challenge it. Follow the part that creates energy.

The goblin can’t make the idea worth publishing.

It can give the idea a rough shape you can finally argue with.

That’s when your standards return. That’s when the writing starts becoming yours.

The goblin got the cursor moving. It earns Employee of the Minute.

Use AI to stop inventing and evaluating the same sentence at the same time.

Let the goblin bring you something.

Then decide what it’s worth.

🧉 What idea have you been carrying around because you can’t find the right way into it?

Crafted with love (and AI),

Nick “Regional Director of Goblin Labor Relations” Quick

PS… Once the goblin gives the idea a shape, you still need standards for making it yours. The free Voiceprint Quick-Start Guide helps you document those patterns:

PPS… I publish about putting AI to work with a frequency that suggests poor adult supervision. Subscribe, then send this to the writer currently asking an empty document to produce the correct answer from thin air.

It’s been sitting there for eight months. You’ve reviewed it four times. You almost shipped it twice. The tag says #someday.

This post is about that note, and the advice nobody finished writing. Everyone taught you step one: capture everything. Every app, every guru, every productivity video with a desk tour in it. And they were right, for once. The half-second after a thought arrives is the worst moment to judge it (it hasn’t cooled, you can’t see its shape, you’re grading an egg on its ability to fly), so saving it all is correct. Then the advice just... ends. Step two was never delivered. You’ve been running half a system for years and calling the warehouse the win.

My step one is voice notes. A thought shows up and I record it in whatever it’s wearing: wrong words, half sentences, no pants. (That describes the thought and, most mornings, the operator.) Then I move on with my life. No typing. No tidying. Polish at capture time is just judgment sneaking back in through the wardrobe. Catch the genius ones, the half-formed ones, the embarrassing ones you’d never say in mixed company, the garbage. Cheap, fast, greedy.

The problem is never step one.

The problem is Whitney.

Whitney has 14,000 notes. Obsidian vault. Custom theme. Forty-one plugins, two of which conflict in a way she describes as “a known issue.” A graph view she screenshots for strangers on the internet like it’s an ultrasound. (It’s a hairball. It’s always a hairball. No one in recorded history has looked at a graph view and learned anything except that the owner needs a hobby, preferably with some sunlight involved.)

Whitney has published three things in two years.

Somewhere between her first highlight and her fourth app migration, the pile became the point.

The Pile Is Not The Product

Notes exist to produce output. That’s the whole job description.

What most writers have built instead is a storage unit: climate-controlled, beautifully organized, visited mostly to admire the shelving. (Some of you pay monthly for it, which makes the metaphor uncomfortably literal.)

Creatives have a hoarding problem, and we fixed it by renaming it research. Capturing feels like thinking. It isn’t. It’s the stretching thinking does before it gets dressed. Highlighting a book feels like reading it twice. Linking two notes feels like having an idea. Reorganizing the vault feels like progress, in the same way reorganizing the garage feels like building something. (You moved boxes. The boxes are very impressed.)

Whitney knows this better than anyone. Whitney spent an entire quarter migrating from Notion to Obsidian, then briefly to Capacities because it felt “calmer.” Her most-linked note is a note about her linking system. She refers to this as infrastructure work.

Most note piles hold ten years of thinking and zero years of saying. And the internet keeps not noticing

Your Judgment Didn’t Disappear. It Moved.

Here’s where the capture-everything crowd and I part ways: they think the system ends when the note is filed. The note being filed is the system clearing its throat.

The judgment you suppressed at the mic (correctly, keep suppressing it) belongs at the other end of the pipe. Not at capture, where everything gets a yes, but at publish, where almost nothing should. That’s the gate that matters. That’s step two, the half nobody wrote down. And it’s where you finally get to be a snob.

And you should be an insufferable snob. Your notes are private. Your post is forever, with your name nail-gunned to it.

Skip the gate and there are only two places to land. Ship everything and you’ve built a slop factory with better intentions. Ship nothing and you’re Whitney, three posts deep into two damned years. The gate is the only reason “daily” and “good” can live in the same sentence. Five filters, in the order I run them…

1. It Has An Opponent

A thought worth publishing is arguing with someone.

There’s a “no, see, the part everyone gets wrong is...” humming underneath it. Open your pile, land on a note, and try to name who’d disagree with it. If you can’t, you’re holding an observation, and observations make the kind of post that earns three likes and a pity restack from your most loyal friend.

“AI is changing writing” has no opponent. Everyone nods, nobody argues, it evaporates on contact.

“Most writers are using AI to sound more correct, which is the exact thing making their work worse” has an opponent. Several. Some of them have newsletters. Publish that one.

2. You’ve Already Saved It Three Different Ways

Scroll your captures and watch for the repeats. The same thought, recorded Tuesday, then again Thursday in different words, then a third time as a voice memo you forgot you made.

That circling isn’t redundancy. It’s your subconscious filing the same complaint until somebody finally listens. (It’s a terrible communicator but a surprisingly good metal detector.)

The notes you caught once and never returned to were either complete or they were nothing. The ones you keep re-capturing have something still trapped inside. Those are your publish candidates, and the beautiful part is your pile already ran this analysis for you. It just never told you, because nobody asked it a question. They only ever fed it.

3. It Makes You A Little Nervous To Ship

Some captures come with a faint “am I allowed to say this” wobble the second you imagine them with your name attached.

Not the reckless kind. The kind where you’ve got a real position and you’re not totally sure you could defend it at a dinner party full of people who do this for a living.

That wobble is the sound of an actual opinion. Consensus is comfortable. Comfortable is forgettable. If a take costs you nothing to publish, it’ll earn you nothing once it’s out, and you’ll have spent a Tuesday confirming the consensus like a notary.

4. It Came With Heat

Watch for the captures that still have a temperature when you reread them a week later.

Irritation. Delight. That hot little flash of “oh, that is so wrong.” Secondhand embarrassment on behalf of an entire industry. (My best material runs about 40% spite, and I’ve made peace with that. My therapist calls it fuel-agnostic.)

A thought that went cold in the file and stays cold on reread is probably done. The ones that still have a pulse weeks later are the ones to grow. Emotion is skin in the game, and skin in the game is what separates a post from a press release.

5. It Shows Up With Its Own Example

This is the one the machines can’t fake, because the machine wasn’t in the room.

Reread the capture. Is there a real instance already attached, or do you have to go shopping for one? Load-bearing thoughts arrive pre-attached to the thing that set them off. The client who said the thing. The post that made you wince. The exact moment you spotted the pattern.

If the example’s already sitting in the note, the thought is real and ready. If you have to assemble evidence after the fact to justify shipping it, that’s the tell: you felt something once, and now you’re dressing it up as a thought. (Nothing against a vibe. A vibe just won’t survive contact with a reader.)

The specific is the moat. The generic is the trap. The machines own the generic now, all of it, in bulk, at wholesale prices. Your lived specifics are the one inventory they can’t stock.

The Stuff That Feels Publishable And Isn’t

Your pile is full of these, and that’s fine. The mistake was never catching them. It’s letting them out.

The recap. You’re restating someone else’s point with no friction added. Felt productive to capture. Adds nothing the second you publish it. The internet did not need a fourth summary of that podcast.

The quote shrine. A highlight from a book, lovingly preserved, tagged, and linked to six other highlights. This is the knowledge worker’s signature move: collecting other people’s thinking and filing it under “my ideas.” A quote in your notes is a guest. It doesn’t pay rent, it doesn’t go out under your name, and if your publication is mostly quote shrines, you’re running a museum gift shop for minds that actually shipped.

The vague gesture. “Things are really shifting right now.” In what direction. For whom. Says nothing, costs nothing, means nothing. Fine in a file. Filler on a page.

The to-do in a trench coat. “I should write about onboarding.” That’s a chore that wandered into the wrong room and put on a thinking face. Keep it as a task if you want. Just don’t mistake it for an idea, and for the love of hell, don’t publish it as one.

The Three-Second Test

Before you publish (not before you capture, never before you capture), ask one question:

Would I bother arguing this with a smart person who disagreed?

If yes, it’s got a spine. Grow it, shape it, ship it. If you’d shrug and change the subject, leave it in the pile. Maybe it ripens. Maybe it doesn’t. Either way it stays private until it earns its way out.

One question. Three seconds. It runs faster than the urge to overthink, which is the only speed that matters when you’ve got a few thousand captures and a finite number of weekends.

And this is the part Whitney never got to, because Whitney is currently evaluating Tana. The pile was never the product. It’s the supply chain. The snob at the publish gate is only ever as good as the greedy lunatic at the capture end feeding them options, and the gate is only worth anything if it actually opens. Often.

Because here’s the math the hoarders keep missing: blurt a hundred captures a week (cheap, remember, you’re already doing it) and even a gate that rejects 95 of them hands you a post for every weekday. That’s prolific. Not because you lowered the bar, but because you raised the intake. The slop factories publish everything they generate. You publish the survivors, and the survivors come out swinging.

Let the rejects rust in the yard. Some of them are next month’s best post; you just haven’t come back for the parts yet.

One more thing about the pipe, since I just spent two thousand words yelling about it. You already know my capture end is a phone and a blurt. The middle is the part nobody believes until they watch it IRL: one rambling voice memo going in, a week of published work coming out, fingerprints intact. So I’m gonna run it live in the Word of Mouth workshop, judgment calls out loud, mistakes included. If you’ve ever wondered whether your blurts could carry a publishing schedule, that’s where you find out.

Register For The Live Workshop

Catch everything in your own unpolished words. Ship what has a pulse. The gate between them is your taste. The cadence is what makes you prolific.

Crafted with love (and AI),

Nick “Snob at the Gate” Quick

PS... Skimmed to the bottom? The whole post in one line: catch everything as voice notes, gate ruthlessly, ship daily. The one part prose can’t show (a single memo becoming the week’s content) is the part I run live in the Word of Mouth workshop.

PPS... If this one made you side-eye your notes app, a like or restack helps it find the other hoarders. And Whitney, if you’re reading this: I know you didn’t restack it. I know you clipped it into the vault instead. I even know the tag you used. It was #someday. It’s always #someday.

For months before that, I’ve been telling you the opposite-sounding thing: fill your you-shaped holes. The spots where you went generic instead of specific. Pour in the realest, most particular version of yourself you’ve got.

So which is it? Carve myself out, or pour myself in?

(Reasonable question. I’d be annoyed too.)

Both are true, because the word “I” is pulling double duty, and only one of its two jobs is the leak.

Job one is the “I” that makes the reader stop and size you up. “I think.” “In my experience.” A credential dropped before you’ve earned the right to drop it. The sentence quietly asks a stranger to run a background check on you before they’re allowed to care about the point. That’s the tax. That’s the “I” I told you to cut yesterday.

Job two is the “I” that makes the reader stop and recognize themselves. Not “I procrastinate sometimes” but “I have reorganized my entire desk twice and produced one sentence.” You said “I.” A stranger heard their own experience. That’s a you-shaped hole, filled. That’s the “I” you keep.

So one sentence pulls off both things I’d supposedly been contradicting myself about. It fills a you-shaped hole with something only you’d write, and it dodges the tax, because nobody audits you for the desk confession. They just nod. They’ve been there too.

No contradiction.

The Part Yesterday Left Out

And it matters more than the cutting did.

Yesterday handed you four ways to cut an “I.” The danger baked into that advice is overcorrection. You read “cut the I,” grab the scissors, and start gutting the lines that were quietly doing serious work. The mirror-I’s. The “me too” lines.

Those are exempt. Don’t run the moves on them.

The cutting was only ever aimed at the I’s that make a reader audit you. The I’s that make a reader recognize themselves aren’t a tax. They’re you, the actual you they showed up for. Leave them exactly where they are, and point the scissors at everything else.

Both posts, folded into one line: cut the “I” that makes them look at you, keep the “I” that makes them see themselves.

You were never choosing between getting out of the way and showing up. The best sentence does both in the same breath.

Disappearing and being unmistakably present aren’t opposites. They’re the same skill. You’ve just been keeping two different scorecards for it.

Crafted with love (and AI),

Nick “Tax-Exempt” Quick

PS... Here’s the dirty secret nobody selling “AI writing systems” will tell you: the model is rooting against your weird specifics. It wants you palatable. Forgettable is safer for it than wrong. The free Ink Sync Workshop is how you overrule a machine that’s optimizing for nobody-gets-mad:

Co-Write with AI

Free Workshop

Ink Sync Workshop…

Read more

5 months ago

PPS... Three buttons down there. Like (you agreed and got shy about it). Restack (you agreed loud, in front of your whole feed, bless you). Subscribe (you’re in, no take-backs, I’ve already mentally moved into your inbox). Pick your level of commitment. There’s no wrong answer except scrolling past like we didn’t just have a moment.

“I’ve spent three years figuring out how to publish every day without becoming a hostage to my own content calendar.”

“You can publish every day without becoming a hostage to your own content calendar. I spent three years proving it.”

Same three years. Same content calendar. Same guy.

Read the first one again and watch what your brain does. There’s a stall. A little hitch where you stop absorbing the idea and start running a background check on the author instead. Who’s this guy? Am I enough like him for his three years to apply to my life? Do I care yet?

That stall costs you readers. And you keep installing it in the first sentence, which is the one piece of real estate where you genuinely cannot afford the rent.

I do it too. Not in the openings, usually (I’ve stewed on this principle long enough that the ledes mostly behave themselves now). It’s the other surfaces that catch me. The cold DM that opens with three “I”s like a man introducing himself at gunpoint. The reply that’s secretly about me wearing a question mark as a disguise. The Note that starts “I’ve been thinking.” The tax never disappears. It just relocates to wherever you stopped looking.

Welcome to the first-person tax.

The Toll Booth You Welded Into Your Own Front Door

Every sentence asks the reader to do something. The good ones ask for almost nothing. They hand over an idea and step out of the way.

A first-person sentence at the top of a piece asks for something first. Before anyone can care about your point, they have to settle a question you quietly dropped in their lap: is this person enough like me that their experience transfers to mine?

You turned the first sentence into a cover charge before the reader could tell whether there was a party inside or six people networking under fluorescent lights.

This isn’t about whether “I” is tacky or self-indulgent or whatever the writing-advice people are wringing their hands about this week. (They’ll tell you first person is “brave” and “vulnerable,” usually in a newsletter that opens with a three-paragraph account of their morning pour-over. Let them.) It’s mechanical. You bolted a toll booth into your own doorway and now you’re confused about the foot traffic.

And the cruelest part? The better your idea, the worse it stings. You wrote something genuinely useful, then guarded the entrance with a pop quiz.

The Camera Is Pointed At Your Own Face

You write from inside your own skull. That’s the entire trap.

You’re standing inside the experience, marinating in it, narrating live. The reader is out on the sidewalk in the cold, deciding whether to come in. “I” is the natural word for the person inside the house. It’s the wrong word for the person you’re trying to coax through the door.

Nobody chooses the first-person tax. You inherit it. It’s the default camera angle, the one aimed back at your own face, and no one ever passes you a note that says turn the damn thing around.

So. Consider this your note.

You probably know a Trevor. Trevor publishes a newsletter called Trevor’s Take (the name is always the first warning shot). Every issue opens with Trevor. Every Note opens with “I’ve been thinking a lot lately about...” Trevor’s About page starts at his birth and proceeds chronologically, like a deposition. We’ll keep an eye on Trevor.

Thirty Seconds And A Highlighter

Here’s the check. You run it on anything before it ships, and it takes about as long as a drunk man needs to decide karaoke was always his destiny.

One. Isolate your first three sentences. Just those. The opening is where the tax incurs the most damage, so that’s where you audit first. (For a Note, a subject line, a DM, the “first three sentences” might be the entire thing. Even better. Less to miss.)

Two. Highlight every I, me, my, we, our. Look at the density. If your opening is wearing first-person pronouns like a band tee at a reunion show, that’s your signal.

Three. Ask one question of each highlight: is this pronoun making the reader recognize themselves, or making them recognize me?

That third question is the whole tool. Recognizing themselves is free. Recognizing you is the tax.

Most of your highlights will fail the question. Good. Failing is the point. Now you know exactly where the leak is, and you can patch it with one of four moves.

Four Ways To Stop Charging Admission

Move one: “I” becomes “you.”

Before: I think most people overcomplicate their publishing setup.
After: You’re overcomplicating your publishing setup.

Shorter, ruder, lands on impact. The reader doesn’t have to wonder whether your opinion applies to them. You aimed it at them on contact.

Move two: “I” becomes implied “you” (drop the pronoun entirely).

Before: I’ve found that batching a month of content kills the part worth reading.
After: Batching a month of content kills the part worth reading.

You’re not always trying to point at the reader. Sometimes you just need to get your own face out of the frame so the idea can stand on its own two legs. Cut the pronoun. Keep the truth.

Move three: “I” becomes “we.”

Before: I keep falling for the same bright-shiny-tool trap.
After: We keep falling for the same bright-shiny-tool trap.

This one’s a scalpel, not a butter knife. Used once, “we” pulls the reader into the same boat as you, and the boat feels cozy. Used in every sentence, it starts to sound like a wellness brand that just discovered the word “community” and won’t shut up about it. Ration it.

Move four: keep the “I,” but make it earn the seat.

Before: I spent three years figuring out how to run a daily publication without a team.
After: You can run a daily publication without a team. I spent three years and one minor mental breakdown proving it.

Watch the order. The “you” sentence pulled them in. Now the “I” shows up as a credential, not a customs form. It’s carrying proof the reader actually wanted. That’s a paid-up “I.” Worth every letter, sitting exactly where it belongs (second, after you’ve earned the right to talk about yourself).

When The Toll Is Worth Paying

Here’s where the lazy version of this advice faceplants.

If the rule were just “never start with I,” you’d be deleting some of the best openings ever written. Plenty of pieces open in the first person and detonate anyway. The pronoun was never the villain. The relationship between your “I” and the reader’s “me” is the villain.

So the real rule is the mirror test.

Your “I” earns its seat when the thing right behind it is something the reader has also lived, named so precisely that “I” stops being a window into your life and becomes a mirror they catch themselves in.

Mirror (works): I argue with people in the shower who aren’t there, and I usually win.

The reader finishes that one for you, silently, three words: ...so do I. You said “I.” They heard “us.”

Diary (doesn’t): I switched to a new note-taking app last week.

Nobody finishes that. They just watch you. You’re up on the little stage, they’re in the folding chairs, and the lights are getting awkward for everybody.

Same pronoun. Opposite result. The difference is whether the reader can stand inside the sentence with you, or has to sit there watching you stand in it alone.

One Tax Collector, Eight Door

You don’t open in one place. A full-stack writer has a dozen front doors, and the tax collector is leaning on every single one of them with his arms crossed.

Post and essay openings. The classic. First sentence stuffed with “I,” reader gone before the piece earns a downward scroll.

Substack Notes and short-form social. Brutal, because there’s no second chance. The “I” Note demands the recognition check inside a feed that’s already deciding whether you exist at all. Lead with their problem, not the weather report inside your skull.

LinkedIn hooks. “I’m humbled and proud to announce that I...” has ended more reach in the feed than any algorithm change. Trevor posts these. Trevor is always humbled and proud. Trevor’s engagement is twelve likes, eight of them from other Trevors who didn’t read the post.

Landing and sales pages. The page is about what they get. Open with “I built this because I was tired of...” and the reader has to translate your problem into theirs before they care. Most won’t run the conversion. They’ll just leave the math undone and the tab closed.

The About page. The worst offender, every time, no exceptions. People write it as an autobiography when the reader is asking exactly one question: is this for me? Answer that first. Your origin story can wait three sentences. (Trevor’s About page is still going. We’re up to his gap year.)

Cold DMs. “I came across your work and I wanted to reach out because I...” Three I’s before you’ve handed over one reason to keep reading. They’ve already started typing “no thanks” and deleting the “thanks.”

Subject lines. The most valuable real estate you own, and “I” torches it instantly. The subject line’s whole job is making them feel the email is about them. “I” makes it about you, unopened, forever.

CTAs. Even the ask. “I’d love it if you subscribed” centers your feelings. “You’ll get [the specific thing]” centers their payoff. One begs. One offers. Guess which converts.

The Sixty-Second Shakedown Before You Publish

You don’t need to rewire how you write. You need one pass, every time, before anything goes out the door. Sixty seconds, maybe less once it’s a habit.

One. Find your first point of contact. Opening sentences, headline, subject line, hook, the first line of the DM. Whatever the reader hits first.

Two. Count the first-person pronouns. Out loud is better. Hearing yourself say “I” four times in one paragraph does something a silent read politely refuses to.

Three. Run the question on each. Mirror, or window? Recognizing themselves, or recognizing you?

Four. Convert the windows. You. Implied you. Earned I. Pick the move that fits.

Five. Leave the mirrors alone. If the “I” makes them mutter “me too,” it’s pulling its weight. Stop fiddling with it.

Six. Ship it.

Run this for two weeks and the checklist dissolves into instinct. You start hearing the tax in your head as you write, the way a guitarist hears a bad note a half-second before their finger lands on it. That’s the actual goal. Not a rule you obey forever. An ear you grow.

Trevor never grew the ear. Trevor refreshes his dashboard. Twelve opens. He blames the algorithm, then the niche, then Mercury being in retrograde. Then he opens a new post, types “I,” and leans back like he’s said something. The cursor blinks. It always blinks.

Your reader showed up willing to care. The least you can do is not charge them at the door before you’ve said one thing worth the toll.

Crafted with love (and AI),

Nick “Recovering First-Person Addict” Quick

PS... Your AI collaborator will reinstall the first-person tax every chance it gets (the statistical average is hopelessly in love with the word “I”). The free Ink Sync Workshop walks through teaching it what to do instead:

PPS... If this earned it: subscribe so the next one finds you, restack it so it finds someone else, and send it to the one writer in your life who opens every single thing with themselves. (You already know who.)

Not someday. Not after some $900-a-month enterprise model eats your archive in a windowless data center while a guy named Chester says “creator moat” into a ring light. Now.

Your rhythm, your sentence habits, your favorite little verbal limp, the specific texture you spent years sanding into shape until readers could pick you out of a lineup blindfolded, all of it can become a prompt, a markdown file, a reusable setting.

A tiny personality costume for the clanker to wear while it tap-dances across the feed pretending to be you. (Rude, frankly. At least buy me dinner before wearing my sentence structure.)

I know because I built a system that does exactly that.

Which makes this next part inconvenient as hell.

Your voice still matters. It gets you noticed. It gets the thumb to pause. It keeps your writing from sounding like a compliance department discovered transcendental meditation.

But it will not make you uncopyable.

Not anymore.

The Receipt Has My Name on It

I’ve spent a year telling creators their voice matters.

A hundred-plus posts. A full training course with multiple version updates. A whole private vocabulary I made up in public and then watched people start repeating back at me, which is both flattering and how cult documentaries begin.

Voiceprint. You-shaped holes. Ensloppification. Throughlines. The kind of terminology that starts as “useful shorthand” and ends with me staring at a Notion page at 1:17 a.m. wondering if I accidentally invented a religion for burnt-out writers.

All of it aimed at one promise: find your voice, protect your voice, teach the machine your voice, and keep the slopgoblins from crawling through the vents and wearing your face.

That promise still has teeth. But it is incomplete. And the receipt has my name on it in permanent marker, with the little flourish I use when I’m trying to look confident while making an avoidable mistake.

I built the Voiceprint system to document how a person writes closely enough that AI can echo them back. The rhythm. The tics. The sentence shapes. The little phrase-level fingerprints. The recurring metaphors that should probably be taxed as dependents at this point.

It works. That’s why people pay for it. That’s also the part where the floor gets oddly squishy, because if I can document your voice well enough to reproduce it, so can someone else.

Maybe not as well. Maybe not with the same care. Maybe not with the part that would survive a second read. But close enough for the scroll.

And “close enough” is where a lot of damage hides out.

Your voice, the thing you spent years developing, can now be turned into a file, a prompt, a cheap mask, a weekend project for someone with your archive, a free Sunday, and the moral posture of a raccoon in a dumpster behind a Cheesecake Factory.

I keep a v4.5 of myself in a text file.

Around here, that document has a name: Voiceprint. Outside this little corner of the internet, it probably sounds like the kind of thing a man invents five minutes before losing custody of a whiteboard.

It is exactly what it sounds like: a real document explaining my rhythm, my sentence habits, my preferred metaphors, my banned phrases, my comedy structures, and the acceptable dosage of verbal damage required to make AI sound less like HR found God.

Four and a half versions of my own personality. Patch-noted like software.

“Improved rant density.”

“Reduced accidental corporate phrasing.”

“Fixed bug where Nick used the word ‘tapestry’ and was still allowed near language afterward.”

Try calling that a moat with a straight face.

Voice used to feel like a wall. Now it’s a photocopier setting.

We Mistook “Hard to Copy” for “Impossible to Copy”

A moat is not “a thing people recognize about you.” A moat is the thing competitors still can’t copy after you show them exactly how it works.

That distinction matters because the creator economy has spent several years treating recognizability like defensibility. They are related. They are not married. They don’t share a Costco membership.

For a while, voice looked defensible because sounding like a specific human was genuinely hard. Most people couldn’t do it. Most AI couldn’t do it. Most brand voice guides were just five adjectives standing on each other’s shoulders inside a trench coat.

“Bold.”

“Warm.”

“Clear.”

“Playful.”

“Authoritative.”

Congratulations. You have described every startup, coach, newsletter, candle brand, and suspiciously expensive productivity app currently in existence.

Real voice was harder than that. Real voice had texture. It had damage. It had weirdness. It had the unmistakable signs of a person who had lived through a few things and developed opinions as a side effect.

So voice felt like protection.

Then the models got better. The copying got cheaper. The floor rose. And the thing that used to separate you became the first thing the machine learned.

Your cadence? Copied.

Your frameworks? Copied.

Your favorite rhetorical move where you stack three clean sentences and then sneak in a longer one dragging a Radio Flyer wagon full of grievances? Copied.

Voice didn’t stop mattering. It stopped being scarce.

A moat made of something anyone can manufacture in an afternoon is no longer a moat. It’s a screen door with a Home Depot receipt and a Christ complex.

Voice Still Gets Them to Stop Scrolling

Voice still does real work.

I’m not telling you to throw your voice into the algorithmic soup and start writing like a horoscope that just got a Canva account. Please don’t do that. The internet has enough sentences that sound like a crystal shop learned conflict avoidance.

Voice gets you noticed. It creates friction. It gives the reader a reason to pause before the feed drags them behind the barn and turns them into engagement mulch.

A voice-lobotomized post still dies on contact. You know the kind. It says things like, “Creators must develop authentic content strategies in the evolving AI landscape,” which is technically a sentence in the same way a smooshed protein bar found under a car seat is technically food.

Fine. Edible under specific emergency conditions. Not something you build a life around.

So yes, keep your voice. Teach the machine how you sound. Stop letting it hand you sentences that smell like enterprise software.

But voice is the door. It’s not the house.

And if your whole AI writing strategy is “make this sound like me,” you’re sweeping the doormat while the roof is missing.

Nice doormat. Very you.

Rain still gets in.

The Prompt Bros Are Looking in the Wrong Place

The prompt bros are looking in the wrong place.

Not because the prompt isn’t long enough. People will keep trying, of course. They’ll write 14-page megaprompts with sections like “Core Emotional Frequencies,” “Brand Soul Parameters,” and “Forbidden Energetic Textures.”

Kyle will sell it for $97. He’ll call it God Mode. He’ll include a Notion template. Someone in the comments will mistake a JSON schema for prophecy.

The world will continue spinning because it has no taste.

But your real edge isn’t hiding out in better prompts. It’s beyond your voice.

The machine can copy the wrapper. It still needs you to bring the thing worth wrapping.

The Four Things AI Can Imitate But Not Originate

If voice is the wrapper, these are the four things worth wrapping:

Taste: what you refuse to publish.
Judgment: what you’re willing to stand behind.
Receipts: what proves you were actually there.
Trust: why people know the difference.

Taste is the part that says, “Hell no.” The machine will add. You have to cut. It’ll give you seven bullet points, three caveats, one “to be clear,” and a paragraph using “transformative” with the confidence of a man who owns three blazers. Your job is to kill the technically fine sentence before it sands the soul off the piece.

Judgment is the call you make before the room gives you permission. AI can generate a contrarian angle. So can a drunk uncle with a Facebook account and a suspicious number of opinions about seed oils. The hard part is deciding which take deserves your name before the evidence is tidy.

Receipts are what prove the point came from a life, not a prompt. AI can imitate the way you make a point. It cannot supply the reason you had the point in the first place. No weird client call. No launch that made $37 and emotional damage. No awkward DM. No “I know this because I lived the stupid version.”

Trust is the part the prompt bros never know what to do with. The reply thread. The private note. The reader who opens because it’s you, not because the subject line performed a circus trick. Relationships are slow, annoying, human, and hard to screenshot as a growth hack.

That’s the part AI can’t originate. It can help you publish more, research faster, edit cleaner, repurpose smarter, and move through the publishing stack with less friction.

Good.

Use the clankers.

Just don’t put them in charge of deciding what matters.

The Dangerous Part Is When It Works

The bad AI outputs are easy to spot. You see them immediately.

“This post aims to explore the transformative potential of AI-powered content workflows.”

Delete. Exorcise the demon. Maybe light a candle.

The dangerous outputs are the plausible ones. They sound pretty good. They sound like you. They have your rhythm. They use your favorite metaphors. They even include a parenthetical aside in the right place, the little bastard.

But underneath the voice, there’s nothing only you could have brought.

No taste. No call. No receipt. No relationship.

Just slop with a nice accent.

That’s the one that fools you. Not the garbage. The polished garbage. The garbage that learned your handshake.

Bring the Substance Before You Bring the Machine

You don’t abandon voice. You demote it from the whole strategy to the delivery system.

Before you ask AI to write like you, give it something worth amplifying. Bring your hot take. Bring your red pen. Bring the thing you learned the stupid way. Bring the detail that makes the template start sweating. Bring the part you almost cut because it felt too specific, too personal, too much like something only your readers would understand.

Especially that part.

That’s usually the bit that makes the piece yours.

Then use AI like a publishing team. Not a replacement brain. Not a slop cannon. Not a one-click ghostwriter wearing your hoodie and silently ruining your reputation while you sleep.

Let it research, challenge, edit, repurpose, test headlines, fix structure, check continuity, and help move the work through the publishing stack.

But don’t let it decide what matters. Don’t let it choose the hill. Don’t let it translate your conviction into consensus.

You stay editor-in-chief.

AI gives you ten angles. You pick the one with blood in it. AI expands. You subtract. AI helps the work travel farther. You make sure there’s still a person at the center of the machine.

That’s the correction.

Not less AI. Better chain of command.

Voice still matters. But the full stack matters more now: ideas, research, writing, editing, repurposing, distribution, audience growth, monetization, reader trust.

Let them steal the wrapper.

The wrapper was never the meal.

Voice gets them to the door. Your taste, judgment, receipts, and reader trust are why they stay.

Crafted with love (and AI),

Nick “Slop Inspector General” Quick

PS… If your AI still sounds like ChatGPT wearing a fake mustache and hoping nobody asks follow-up questions, start here:

Co-Write with AI

Start Here

Stop Sounding Like ChatGPT: Build Your First Voiceprint with the VAST Framework…

Read more

5 months ago · 1 like

The Voiceprint Quick-Start Guide helps you give AI a better map of your voice, your rhythm, your banned phrases, and the little verbal damage that makes your writing yours.

Not magic. Not soul cloning. Just a better operating manual for the machines so you can stop receiving drafts that sound like they were swept out from under airport carpet and formatted as advice.

PPS… Like, comment, restack, or send this to your cousin three times removed because statistically someone in your family tree is currently using AI to write “thrilled to announce” posts and must be stopped before they launch their next Etsy shop.

Yesterday I made the case for why Substack Chat earns its keep (retention, not growth, the room where casual readers turn into people who’d feel a loss if they left). Today I stop philosophizing and hand you the playbook. Three plays. The prompts that pull a silent room into conversation, the weekly rhythm you can run without it becoming a second job, and how the room starts talking when you’re not even there.

Why Your Openers Keep Getting Crickets

Most Chat prompts get silence for the same three reasons, every time.

They’re too broad. “What’s on your mind this week?” is not a question. It’s a blank page with a cursor blinking at someone who came to Substack specifically to escape blank pages.

They sound like a substitute teacher. “Hey everyone!” is what you say when you’re addressing a room you can’t name a single person in, cranking the cheer to cover for not belonging there yet.

And they ask for labor. Anything that needs a reader to compose a paragraph, dig up a link, or actually think hard gets filed under “later”… And “later” always means “never.”

A good Chat prompt is answerable in one line, on the spot, by someone holding their phone in a grocery line. You’re not asking for an essay. You’re handing them a slot to fill.

(Worth saying out loud, because people blur these constantly: Chat is not Notes. Notes are a public broadcast aimed at strangers in the feed. Chat is the room your subscribers already walked into. You can be more insider, more familiar, more “you already know me” in here. Different surface, different voice.)

Here are openers you could post today. Notice they all give the reader an easy thing to grab, and notice that you go first every time. The room mirrors the host. If you want replies, model the reply.

• “Quick gut check. Morning writer, or ‘one more coffee and suddenly it’s 4pm’ writer? I’m aggressively the second one.”

• “Finish this. The one AI phrase that makes me want to launch my laptop across the room is ______.”

• “I’ll start. 212 articles saved to ‘read later’ that I know I’ll never read. What’s your number? Just the number.”

• “Drop one Substack you read that nobody’s talking about yet. I’ll leave a comment on each of their latest posts.”

• “One word for how your writing week went. I’ll go first: feral.”

The “drop one Substack” move does something sneaky and good. It makes the reader the expert for a second, which feels great, which is exactly the feeling you want them associating with your room.

Then reply to every single person who bites. Not “love this.” React to a specific line, extend their point, complicate it gently, or ask one narrow follow-up. (Those are the same four moves that win you a stranger’s attention in the comments, which I broke down in Good Writing Doesn’t Get Found. People Find People. The room is just where you run them on people who already chose you.)

A Rhythm You Can Run In The Grocery Line

The two ways to wreck a Chat are equal and opposite. Treat it like a second newsletter and you burn out by week three. Treat it like a pitch channel and the room mutes you by week two. Same self-inflicted bullshit, different costume.

The fix is daily presence, not daily posting. Those are two different jobs.

Here’s a week a solo writer can actually keep.

Early week, drop one conversation starter from the list above. Two minutes.

Midweek, give them access to something. A screenshot of your mess. A line you cut and why. A thing that broke. No question required. People respond to being let into the room behind the room.

End of week, run your ritual (more on that in a second).

And every day in between, the only real work: scan what came in and reply. Same day if you can. This is five minutes, not fifty.

Put it on a rhythm so you’re not relitigating “should I post today” every single morning. I’m a broken record about building writing habits on consistency instead of inspiration, because inspiration is a flake and consistency shows up even hungover (the whole case for ritual over willpower is 30 Minutes, Same Prompt, Every Sunday).

One discipline holds the entire thing together. Give roughly eight to ten times before you ask for anything once. If you run three threads a week, that’s nearly a month of pure giving before you’ve earned a single “by the way, paid subscribers get the templates.” Most people invert that ratio and then act surprised when the room treats them like a guy handing out flyers at a silent meditation retreat.

The Day You Go Quiet And It Doesn’t

Everything so far still has you holding the room up. The goal is a room that holds itself up. That starts the moment readers begin talking to each other instead of only to you.

You engineer it on purpose.

Introduce people. When two readers are circling the same problem, say so out loud. “Devon, you need to see what Maria said about this last week.” You become the connective tissue, then you step out of the middle of the conversation you just started.

Name your regulars. The ones who show up every week, call them by name, reference what they said last time. Recognition is the easiest yet strongest retention move there is. Someone who knows your name is in orbit. Someone whose name you know back is on their way to the center.

Run one recurring thread. Same prompt, same day, every week. “Friday confession.” “What broke this week.” Rituals do the work random posting can’t fake, because the room starts showing up before you post. They know it’s coming. That’s when the thing develops its own pulse.

Then comes the step-back. Once the regulars are talking to each other, you climb down out of the lifeguard chair. You post less and react more. You let a question sit for an hour without your answer, because someone in the room will answer it first. They always do, once they stop treating the room like it’s yours and start treating it like theirs.

That progression has a shape. Close (they know your name) becomes core (genuinely invested, they buy, they open up) becomes the people who’d notice you were gone. The ones who’d reply to a missed week with “everything okay over there?” That’s not a number on your dashboard. It’s the entire reason the room exists.

It’s not that the room runs without you. It’s that you’ve baked enough of yourself into it that it keeps its own heartbeat between your visits.

A subscriber list is a stack of permission slips. A Chat that talks back is a place people belong. The work is small, daily, and deeply unglamorous, and it’s the closest thing to a moat a solo writer on Substack ever gets.

Crafted with love (and AI),

Nick “Off-Duty Chat Lifeguard” Quick

PS... got a writer friend whose Chat tab has tumbleweeds rolling through it? Forward this. They’ll either thank you or quietly resent you. Both count as engagement.

PPS... I’ve been quietly helping a few clients jump-start their Substacks, and I packed what actually works into one guide. Want your first 500 subscribers, or just want to bolt 500 more onto what you’ve already got? Hit reply (or flag it in the comments) and it’s yours. No funnel. I’m too tired to build a damned funnel right now.

Substack handed you the tool for that, and you’re probably running it as a second mailbox.

The obvious part first, so we can move past it: Chat won’t win you a single new subscriber. It’s private, subscriber-only, invisible to anyone who hasn’t already joined, and no stranger will ever wander in. Everybody who’s opened one figured that out inside a week, decided it was pointless, and let it go dark in a corner next to their 3rd abandoned podcast.

It wasn’t pointless. It was the one place you build what the writing can’t: readers who’d have to walk out on people, not just close a tab.

The public work (Notes, comments, restacks) earns you the subscriber. Chat is where you take it from there, walking that reader inward: warm to closer, closer to the inner ring, your best people kept from quietly drifting out the back while you weren’t looking. The inner orbit (the warm ones, the readers already worth something) holds your loyalty and your money both. And every one of those relationships is one neglectful month away from being worth nothing at all. Chat’s the best tool Substack hands you for keeping them warm.

The mechanism’s worth understanding instead of taking on faith.

A reader who only gets your posts is subscribed to your content. The week your writing has an off stretch (it will, you’re not a machine, thank god) the whole math of staying is “was the work good enough this week.” But a reader who’s posted in your Chat, gotten a real reply, had their name said out loud, traded jokes with two other members they now half-know? That person’s subscribed to something way stickier than your prose. Leaving stops feeling like unsubscribing and starts feeling like bailing on people who’d notice they were gone.

Nobody does that over one slow month. Belonging’s a switching cost, and it’s the strongest one you’ll ever build on purpose.

Now the part nobody warns you about until they’ve already torched it.

Every new thread can fire a push notification to every subscriber. And you can staple an email onto any thread with one little toggle, dropping it straight into every inbox you’ve got. That toggle is the single most abusable button in the whole feature. Treat your Chat like a content feed that needs feeding (six threads a day, half of them emailed) and you won’t build a community, you’ll train people to mute you. And a mute isn’t contained. It frays the notification line you rely on for everything else, and some of those muted folks quietly stop opening the posts too. Congratulations, you’ve automated your own slow decline.

So the email toggle is for events. The live session. The actual announcement. The flagship thread of the month. Not the fucking Tuesday check-in.

And the flip side, the thing that sends new hosts spiraling for no reason: silence is normal. A Chat isn’t a Discord server with a caffeine problem. It’s episodic. It’ll sit dead quiet for three days and then go off around the right prompt on the fourth. A six-hour gap isn’t a verdict on your worth as a human. Post it, tend it, let it breathe, and stop refreshing the room every twenty minutes like a guy waiting on a text back from someone who’s clearly just not into him.

One gut-check before you ever switch it on: am I doing this to pull people in, or to hold people close? If the honest answer’s “pull in,” put the mouse down. You don’t need a room. You need to go get found first, because a room can’t do that for you.

But if you’ve already got readers in your orbit (people who’d answer if you actually asked) then you’re holding the raw material a room runs on. That’s who Chat’s for. Not the strangers you haven’t met yet. The people already standing close, waiting for one good reason to come closer.

Notice I haven’t actually told you how to do any of this yet. That’s tomorrow. Check your inbox in the morning for the real tactics: the prompts that pull a silent room into a conversation, the weekly rhythm that keeps it going, and the moves that turn a subscriber list into a room that runs on its own. Today was the why. Tomorrow’s the how.

Getting found pulls strangers in. Chat holds the ones you’ve already got close. Run one when you needed the other and it fails on you quietly, which is the most expensive way anything ever breaks.

Crafted with love (and AI),

Nick “Keeper of the Inner Orbit” Quick

PS... If you got something out of this, two small asks. Subscribe if you haven’t (the button’s right there, doing its little job). Then forward it to the writer you know who flipped on Chat eight months ago, posted “hey everyone!” into the silence once, and hasn’t been back since. They need this more than you do. They won’t admit that either.

PPS... Most of what I just described, I run through an engagement platform I’ve been building for clients on the hush-hush, code-named Stackwise. It’s the part of today’s post that stopped fitting in my head: who’s drifting, who’s close, who needs a nudge this week. Curious what’s going into it? Reply to this email or flag it in the comments and the First-500-Subscribers guide is yours. (It’s a deeply unsexy opt-in. I just have a soft spot for the people who bother to hit reply.)

Geoff has 200 subscribers. He’s had roughly 200 subscribers for two years.

He’s decided this is because readers got dumber. (Narrator: readers did NOT get dumber.)

Here’s what Geoff believes, with the serene confidence of a man who has never once been wrong out loud: the work speaks for itself. Write well enough, long enough, and the audience finds you. So Geoff writes well. He does not comment. He does not restack. He has never left a reply on another living person’s post, because engagement is for writers who can’t write. He publishes his beautiful thing into the feed, refreshes the stats, and waits for the world to notice the obvious. (I would like to find whoever coined “the work speaks for itself” and ask them to name one time the work picked up a phone. The work has never spoken. The work sits in a drawer being brilliant while you eat dinner alone.)

The world, meanwhile, has 4,000 unread emails and a dentist appointment. The world is not coming to lend a helping hand.

Three notches below Geoff on raw talent (clumsier sentences, fine ideas, an author photo that’s just her squinting into her front-facing camera in a disorganized kitchen) sits a writer with 40,000 subscribers and a paid tier that covers her rent. Same niche. Same two years. She did exactly one thing Geoff refuses to do.

That one thing is the whole post. And you can start doing it today.

The Algorithm Reads Signals, Not Sentences

The ways you actually get found on Substack come down to a handful: a recommendation (one publication pointing its readers at you), a restack from someone with real reach, a mention from a name people already trust. Those are the channels that move real numbers, and they last. Not one of them is a prize the algorithm hands out for writing well.

Every one of them is a human being deciding to spend their own credibility on you. (Credibility they spent years building and are now lending to another, for reasons that’d better be good.)

This is the part everyone gets backwards about the algorithm, so let me be unkind about it. The feed doesn’t push “good.” It doesn’t read nor does it give a damn. It pushes what’s already moving. It watches who replied, who restacked, who carried your thing into their own house, and it amplifies the motion that already exists. No motion, no amplification. You can write the sharpest post in your corner of the platform, hit publish, and watch it land softly on the couple hundred people who already subscribed before sinking into the silt at the bottom of the feed by the weekend. It just sits there. Excellent. Ignored. The feed moves on, because it never knew the post was good. It only knew that nobody moved.

We built an entire platform for writers and then quietly arranged it so the writing matters least and the showing-up matters most. Nobody puts that on their landing page.

You probably grew somewhere else before this. Somewhere the machine could decide to make you famous on a Tuesday for no reason at all. Not here. Here the machine waits to see if anyone cares first, and “anyone” is people. People Geoff has decided are beneath him. (Geoff refreshes his stats. Flat again. He concludes the algorithm holds a personal grudge. The algorithm doesn’t know Geoff exists. That is, in fact, the entire problem.)

How This Grows Your List, Not Theirs

Here’s the connection nobody draws for you, and it’s the reason none of this is charity.

A comment is writing. It’s a few sentences of your actual thinking, published. The only difference between your post and your comment is the street it’s standing on. Your post stands in your own house, where everyone reading it already subscribed (you are performing for people who bought tickets months ago). Your comment stands in someone else’s house, in a room packed with people who have never heard of you and are, this very second, paying attention to something.

That’s the move. A sharp comment is a free sample of your writing, set down in front of a borrowed audience. It is the one thing your own posts physically cannot do, because your own posts only ever reach the people you already have. (Geoff’s posts reach Geoff’s 200. They will reach Geoff’s 200 in the year 2050, assuming Geoff is still refreshing.)

Now follow one comment forward, because it pays you twice.

The first payoff is today. A few of the strangers in that room read your two sharp lines, think who is this, and tap your name. Some land on your publication and subscribe. Those are readers you couldn’t have reached from your own post in a hundred years, because your post never went anywhere near them. Ten good comments a day, in rooms full of the right strangers, is ten samples of your writing in front of new people, every single day. This is distribution. The exact thing your posts were failing at. It was hiding in the comment box the whole time.

The second payoff is slower and much bigger. The writer whose room you keep showing up in starts to recognize your name. You stop being a stranger and quietly become “oh, that one, the one who actually gets it.” And one day, with no prompting and no pitch and no awkward message, they restack you, or mention you, or recommend you to their entire list. That’s the big door. The one that sends 400 readers in an afternoon. It doesn’t open by magic. It opens because you spent two months being genuinely useful in someone’s comments with your hand nowhere near your own pocket. And it keeps paying, because a recommendation doesn’t spike and vanish, it sits on their page sending you readers for months off work you did one time. (Try getting that out of a post. Posts have the shelf life of a gas station hoagie.)

Both payoffs trace back to the same small act. So: same niche, same talent, same hours at the keyboard. The woman at 40,000 let her writing leave the building. Geoff guards his at the door like a bouncer who hates money.

Stop Clapping

I spent months doing this wrong, so I’ve earned the right to be smug about it now.

My comments used to be applause. Dressed-up applause, but applause. I never typed “Great post,” because I was better than that. (I thought I was better than that.) I’d leave four articulate sentences that, wrung of all their air, came out to “So true” and “Needed this.” A standing ovation with footnotes. Hundreds of them. Felt productive. Grew by approximately nobody. (I write about making AI sound human for a living. I was leaving comments that sounded like a microwave beeping. The irony was available to me the entire time. I ignored it for an embarrassingly long stretch.)

Here’s the problem with applause. Take “Great breakdown, so important!” and hold it under a light.

It anchors to nothing. (The writer can’t tell whether you read the post or fed it through a parser.) It makes no point. (You have contributed the conversational value of a bobblehead.) It would copy-paste onto ten thousand other posts without changing one syllable. (Which tells the writer precisely how much of you is in it.) And it ends on an exclamation mark doing the job three real words should have done. (Punctuation standing in for having had an actual thought.)

The writer glances at it, feels the faint warmth of being noticed by something, and goes back to their afternoon. Nobody tapped your name. Why would they? A wave is not a sample of anything.

The comment that works makes a writer stop, read it twice, and click your name to find out who the hell you are. Watch the gap.

Dead: “Great breakdown, so important!”

Alive: “The ‘six months of expenses’ rule is something everybody repeats and nobody pressure-tests. For someone with lumpy freelance income that’s a completely different number than for someone salaried, and the advice never admits it.”

Same post. Same sixty seconds of your life. One earns a heart and reaches the silt. The other earns a reply and a fistful of profile clicks, because it does three things the wave never does.

It anchors to a specific line. Not the post in general. The actual sentence. Naming the exact line is proof you read the thing, and a small flattery the writer feels in their spine, because you noticed the part they sweated over while everyone else skimmed.

It makes one real move on that line. One. (Not five. Five is some dude explaining your own essay back to you at a party.) The move is the entire value of the comment, and it’s what almost everyone skips, which is exactly why almost every comment lands with a wet thud.

It takes nothing out of the room. No link. No “I wrote about this too.” No “thoughts?” tacked on the end, because you are the guest, and the guest does not start rearranging the host’s furniture. You land the move and you leave. The give is the whole thing.

Four Moves You Can Use Today

A comment that lands does one of these four things, not all of them at once. Pick the one the post actually calls for, do it in two lines, get out.

React to a line. Pull one specific sentence out of the post and respond to that, not to the post in general. “This line stopped me: ‘most advice is just survivorship bias in a nicer outfit.’ Which means the only advice worth a damn comes from the people who also lost, and nobody ever interviews them.” The writer feels seen by someone whose attention seems worth having. (That’s the hook. People will forgive almost anything except not being noticed.)

Extend it. Take the idea one step further with a concrete case the writer didn’t mention. “This holds for pork chops too, which nobody talks about. Salt them the night before and a $4 chop eats like a steakhouse the next day.” You didn’t agree. You added a room to their house, and adding a room is what equals do.

Complicate it. Introduce a real tension the writer skipped. Not contrarian for sport (everyone can smell the guy who disagrees as a personality trait). The genuine edge case. “Works until the feedback runs upward. A junior telling a VP the plan is incoherent pays a career cost the framework never prices in.” Taking someone’s idea seriously enough to find where it cracks is a compliment most writers get roughly never.

Ask one narrow question. Not a vague one. A question so specific they can answer it in a sentence and actually want to. “Did the bigger-buffer version change how your readers behaved, or did they nod and keep the same number?” A sharp question hands the conversation back, which is how a one-off becomes a back-and-forth, which is how a stranger ends up knowing your name.

Notice not one of these asks you to be clever about yourself. You’re being useful about them, in public, where the writer and every lurker in the thread can see exactly how your head works. That’s the sample, doing its job while you do nothing else.

Leave Ten. Take Nothing.

Pick a number you can hit every day without negotiating with yourself. Ten is a good target. Five is fine. Three on a day when everything’s on fire still beats the zero you’d otherwise leave. The count was never the magic. The magic is that you keep showing up in other people’s threads with something real, on the days you feel like it and the days you don’t, which is the exact thingy nobody else can be bothered to do.

Each comment is a sample of your writing standing in a room you don’t own. More rooms, better odds a stranger taps your name or a writer starts to recognize it. But the second you’re padding to hit a number, you’re clapping again, and a small niche smells that from across the building. (That's Geoff's whole method. Volume, vocabulary, and the quiet confidence of a man talking to an empty room.) So the rule isn’t a quota. It’s that every one you leave is real, because a generic comment doesn’t merely waste your time. It tells the whole room you’ve gone on autopilot.

Aim them at people whose audiences you’d actually want to borrow. Peers in your lane, the occasional writer a notch bigger than you. Not whatever floats past your thumb on the scroll. To find them, start with the publications you already read, then watch who they restack and who keeps showing up in their comments (those people are standing in the exact room you’re trying to slip into). Work the Notes feed for your niche too. A sharp reply on a busy Note gets seen faster than a comment on a week-old post, and can put you in front of more strangers in an hour than that post will all month.

Two rules keep them honest. Take nothing out of the room (the link, the “check out my piece,” the soft little mention of what you sell, all of it goes somewhere else, never here, because the second you reach for the audience you stop earning it). And sound like yourself, not a language model wearing your name, because the sample a generic comment leaves is a sample of a machine, and machines don’t get recommended.

Send One. Want Nothing.

Once a day, on top of the ten comments, send one message.

Not a pitch. Not “loved your post, here’s a link to mine.” One genuine, personal note to one real person, asking what they’re working on or offering something you can actually help with. No ask attached. No hook waiting in paragraph three. A hand extended on purpose, to someone whose work you’ve been paying attention to.

Here’s why that single message might be the highest-yield thing you do all day.

Almost every message a creator gets wants something. A pitch, a swap, a favor, a “quick question” that is never quick and never one question. Every one of those hands a person doing the cross-legged bathroom dance another tall glass of water. Yours is the one that points at the door, before you’ve asked for a thing, which makes you the rare name in there they actually want to hear from again.

The inbox is where the real moves get made. A comment is public and shallow by design (it can only do so much in a stranger’s thread). A message is private and direct, and the recommendation, the cross-post, the collaboration, the “want to come on my podcast” all get organized in a one-on-one conversation, never in a comment section. The help-first message is how that conversation starts with no agenda, which is exactly what gives it the standing to grow one later.

The math is quietly ridiculous in your favor. One real message a day is thirty a month, a few hundred a year, each one an actual human you reached one to one with your hand nowhere near your pocket. Most won’t reply. Fine. The few who do are how you build the small stack of relationships that ends up carrying the entire publication.

One catch, because “how can I help” rots fast into its own kind of pitch. Mean it. Name their actual work, offer the specific thing (the introduction, the resource, the answer to the problem they posted about on Tuesday), not a limp “let me know if I can ever help,” which every recipient correctly hears as a handshake right before a sales pitch. If the offer isn’t real, don’t send it.

What that actually looks like, stripped of theory: “Your piece on pricing for freelancers said the thing I’ve been fumbling toward for two years, especially the part about hourly rates quietly training clients to nickel-and-dime you. Unrelated, you mentioned you’re redoing your sales page. I’ve rebuilt a few that started converting, and there was one change that did most of the work. Want me to send you the before-and-after? No pitch, I just got nerd-sniped by the problem.” Three sentences. It names their actual work, offers one specific real thing, and asks for nothing. That’s the entire shape.

Geoff would rather lose a toe than send a message that doesn’t serve Geoff. So Geoff doesn’t send messages. His (lack of) growth has noticed.

That’s the whole practice. The next post you read in your niche, don’t applaud. Find the one line carrying the weight, say one true sentence about it, and leave without taking a thing. Then send one message to one person, and ask how you can help.

Geoff won’t do any of this. Geoff is currently posting a thread to his remaining subscribers titled “Has Substack Lost Its Soul?” and refreshing it for replies. There are none. There are never going to be any. The room is empty because Geoff has never once walked into anyone else’s.

You’re not Geoff. You’ve got ten minutes, a comment box, and a message to send.

Your writing was never the bottleneck. The silence around it was. And silence is the one growth problem you can start fixing today: ten real comments in other people’s houses, and one message that wants nothing back.

Crafted with love (and AI),

Nick “People Finding People” Quick

PS... Two things, in ascending order of effort. One, subscribe if you haven’t (the button does the rest). Two, forward this to the person in your life who keeps insisting they “don’t get Substack.” Your mom counts. So does the friend who’s been about to start a newsletter since the Obama administration.

PPS... I’ve been building an engagement platform behind the scenes, code-named Stackwise. It’s most of this post turned into a system I run my own operation on. Want to know what I’m wiring into it? Hit reply, or say so in the comments, and I’ll send you my First-500-Subscribers guide. (Not the most glorious opt-in ever devised. I just feel like rewarding the people who actually reply.)

PPPS... Geoff made it this far and left without commenting. He’s letting the work speak for itself. We may not hear from Geoff again.

Three years of creators arguing about whether the detectors work. Whether the seams show. Whether anyone besides them actually cares. (Spoiler from someone who’s personally cared about all of this for at least three years: not really, occasionally, almost no one, and the third answer makes the first two irrelevant.)

Three years of energy spent on a question that was rigged from the start, by an audience that was never asking it. (Worth saying out loud at least once. Three years. Down the crapper.)

I’ll be blunt: detection-proofing is the most embarrassing vanity project the creator economy’s produced this decade. And we’ve produced some real shitty ones.

It feels like quality control. It produces fuck-all. You’re polishing the lock on a door nobody gives a shit about opening, while the back wall of your house is missing, the readers are slowly wandering out through the gap, and you’re very carefully reading the engraving on the doorknob to make sure it still says your name.

Detection chases a moving target. Every model release makes the detector worse by mathematical inevitability. The creators “winning” detection this week are passing tools that already aged out yesterday afternoon. (And your audience couldn’t tell either way, which is the part nobody likes to admit out loud at the parties they’re starting to not get invited to.)

There’s an entire cottage industry of detection apps, all of them obsolete by the time you finish reading their landing page. (Excellent business model for them. Less excellent for you, the customer, who paid for a tool that was wrong about its core function before you opened the email confirming your subscription. Capitalism’s something else.)

So If The Audience Never Cared About Detection, What The Hell Did They Care About?

Whether you can defend the argument. That’s the whole shape of it.

Not whether AI helped. Not whether the seams show. Not whether your sentences pass some statistical sniff test that was rigged by the same companies producing the slop the test is supposed to catch.

Whether, when a sharp reader replies with “wait, what about X?” you can answer from your own thinking. Or whether you have to run back to the chat window like a child who forgot his lines, paste in the reply, and ask the model to please help you respond convincingly to the reply about the post the model wrote in the first place.

Pull up your last post. Imagine the reply. Could you defend the central claim without the model in the room?

If the answer’s no, the argument isn’t yours yet. It’s borrowed shape. Which is fine for brainstorming. Less fine for a published piece with your name on the masthead and your face as the avatar.

That’s question one. There are three more, and they get progressively more uncomfortable.

The Audience Doesn’t Audit. It Drifts.

Readers don’t audit your posts. They never will.

Nobody’s at their kitchen table running your newsletter through detection software. They have lives. (Or at least the kind of approximation of lives that lets them clear their inbox and feed the dog before bed.) The whole detection panic’s happening exclusively inside the creator’s skull, the way a forest fire happens without consulting the trees about scheduling.

What the audience IS doing, slowly, over time, is noticing.

They notice who can defend their arguments. They notice who has a real position and who has a model’s polite paraphrase of one. They notice who writes with the confidence of someone who’s actually thought about the thing, versus the confidence of someone who’s been told what to think by something that can’t, by definition, think.

This noticing doesn’t show up as a banhammer. There’s no callout post. No moment of public reckoning. Just drift.

Less citation. Fewer recommendations. Slower growth. Numbers that used to climb start sitting still, then start dipping, and nobody can quite explain it. (You’ll explain it to yourself by blaming the algorithm. The algorithm’s also blaming you. Neither one of you is wrong, which is somehow worse than if one of you were.)

Here’s question two: Can you defend why the piece is shaped the way it is?

Why this opening and not the two others you tried. (And there should be at least two others. If you only had one opening, you didn’t write a piece. You shipped the model’s first draft and called it yours.) Why this order. Why this example over the three alternatives.

If you can’t defend your structural choices, you didn’t make them. The model made them and you signed off. (There’s a kind of person who joins a cult and spends ten years insisting it was his idea. The robe was a gift. The compound was a sabbatical. The shaved head is just easier in summer.)

The Fingerprints Problem

Detection-proofing fails because it solves the wrong problem.

The audience isn’t asking “did AI touch this?” They’re asking “is there anyone home?”

You can produce a piece that’s 100% AI-written and have it feel inhabited, because you pushed back, edited hard, deleted the model’s best paragraph for the third time because you knew it was wrong, and forced your own structural choices through like a small-town mayor refusing to budge on the parade route. You can produce a piece that’s 100% human-typed and have it feel hollow, because you wrote it on autopilot during a 6 AM Zoom with the camera off and your soul on screensaver.

Fingerprints come from thinking. The typing’s incidental.

Which brings us to question three: Where did you overrule the model?

If you never pushed back, you didn’t co-write. You transposed. A real edit (a deletion, a reorder, a flat “no, try again, do it differently, you’re embarrassing both of us”) is one of the cleanest signs that actual thinking happened. The presence of friction proves two minds were in the room.

The absence of friction means you were alone with a very fast typist who can’t tell you when you’re wrong, doesn’t care if you’re wrong, and will produce confident wrongness all day long because that’s the job and it’s good at it.

If your process is “prompt, paste, publish,” you’re not writing anymore. You’re signing forgeries. The model paints. You add the signature. The signature is the only part of the transaction that’s actually yours, and even that is starting to feel borrowed. (The editor who would have recommended you can spot a forgery at fifteen feet. The conference programmer is now booking the writer who can defend her sentences in a Q&A. The reader who used to forward your stuff to her smart friend has stopped, because forgeries don’t travel well. People share originals. Drift is silent. But it has sharp teeth.)

The Sentence Only You Would Have Written

The last question’s the hardest. (It’s also the one creators most consistently fail, including myself. The first time I ran my own work through this check, I learned things about my writing I’d have preferred not to learn, and then I learned the same things again about the next piece, and the next, and at some point I stopped being surprised and started being useful to myself.) It’s also the one that separates work that compounds from work that dissolves like sugar in rain.

Is there a sentence in this piece that only exists because YOU did the thinking?

The kind of sentence that emerged from sitting with the problem long enough to find a specific thing nobody else was going to say. The kind that wouldn’t surface from any prompt because no prompt would know to ask. The kind you couldn’t outsource if you tried, because you didn’t even know it was coming until your hand typed it and you looked at it and thought “huh, okay, that’s mine.”

If you can point to that sentence, the piece is yours. The byline is honest. The work is real.

If you can’t, what you’ve published is a remix squatting under your byline, and every reader who sticks around long enough to read three or four of your posts in a row will start to feel something they can’t name. They’ll move on. Not all at once. Slowly. The way people stop calling a friend who only ever talks about himself.

The recap, since this got dark

Four questions. Run them on every draft. Not as a purity test. As a calibration.

1. Could you defend the central claim without your AI in the room?

2. Can you defend the shape of the piece?

3. Where did you overrule the model?

4. Is there a sentence only you could have written?

If a draft fails one of these, the fix isn’t to rewrite the surface. The surface is where detection-proofing happens. The surface is also where ensloppification happens. Both problems live in the same corner of the workflow, and both come from the same source: not doing the fucking thinking the model bypassed.

You can keep polishing the lock. The windows are still open. The readers are still drifting out the back wall. (The back wall’s been missing for approximately three years, but you’ve been busy.) The audience is still building its quiet, unspoken standard, and you’re still optimizing for the wrong audience entirely. Detection theater’s collapsing in slow motion, and the creators still anxiously running their drafts through tools will wake up in a year wondering why their numbers stopped moving.

The creators running the defensibility check will still be there. They’ll be tired. They’ll be cranky. (They’ll have published five drafts they almost didn’t because they failed question four three times each. This is the work. The work’s occasionally insufferable. The alternative’s worse.) They’ll be writing something nobody else could’ve written.

You can join them. Or you can keep signing forgeries.

Crafted with love (and AI),

Nick “Patron Saint of Lost Vanity Projects” Quick

PS... The four-question check is the doorway. The Voiceprint Quick-Start Guide’s the foundation underneath it. (You can’t write a sentence only YOU would write until you’ve done the work of figuring out what “only you” sounds like. The guide’s where that part starts.) Grab it:

Co-Write with AI

Start Here

Stop Sounding Like ChatGPT: Build Your First Voiceprint with the VAST Framework…

Read more

5 months ago · 1 like · Nick Quick

PPS... Co-Write with AI publishes pretty much every single day, which is partly discipline and partly the only thing keeping me from reorganizing my pantry alphabetically by spice family at 3 AM. Forward this to a creator who’s still posting detection scores like they’re meaningful currency. They’ll either thank you in six months or unfollow you immediately. Both outcomes count as modest improvements.

Same week. Two organizations. Two policies that look opposite if you tilt your head and squint.

(They’re not opposite. Same direction, viewed from two ends of a long hallway. Bear with me.)

Both companies wrote down what AI was and was not allowed to do, in operational language a child could enforce. The Times put it in a memo. TikTok put it in code. Both did it.

The rest of us said “human in the loop” a hundred times this year and called it a policy.

(The way a kid in a Halloween costume tells you he’s a doctor. Very confidently. While holding a plastic stethoscope. Diagnoses entirely vibes-based.)

The Subtle Shift Nobody Sent You a Memo About

For two years the AI debate was about whether using it made you a phony. Should you let it write the post. Should you let it ghostwrite the email. Should you let it touch your work at all, or should you do the noble thing and spend Saturday morning grinding out a Substack essay by hand like a Victorian governess polishing a candlestick.

That debate ended a while ago.

The answer was yes, most of you do, all of you will, and the moralizing has the energy of a parent forbidding booze at a frat house six hours into a raging kegger.

(I was one of the moralizers. About six months, all told. It is embarrassing in retrospect, which is the only reason I’m putting it in writing. Free your Saturday. Skip my mistakes.)

While everyone argued about whether using AI was cheating, AI secretly took the next floor of the building. The campaign layer. The publish-button layer. The layer where decisions used to require a coffee and an argument. TikTok’s MCP server is one example. Picsart’s agent marketplace was another. Adobe rolled out custom Firefly models last quarter. Cloudflare predicts bots will outnumber humans online by 2027, which is the sort of statistic you read once and decide you will think about later, when you have more bandwidth, or possibly a stronger drink.

(I think about it constantly. It is not a fun statistic. There is no fun statistic in this paragraph. Moving on.)

AI is not your assistant anymore. It is the operator of the systems you used to run yourself, and most of you haven’t noticed, because the operator is polite, prompt, and doesn’t ask for raises.

Which means the skill that mattered most for the last two years just got demoted while you were looking the other way.

The Dead Question (May It Rest In Peace)

“Should creators use AI?” is a dead question. We should let it lie in the road like the small dead possum nobody is willing to bury.

The question was always shaped wrong. It treated AI like a moral test instead of a tool, which is the kind of category error you make when you have never actually used the thing you are moralizing about. The honest answer was always “some of us, in some ways, depending on the work.” Now even that answer is irrelevant, because the option in front of you is no longer “AI yes” or “AI no.”

It is “AI you noticed” or “AI you didn’t.”

(Most creators have not priced this into the conversation. They’re still trying to decide whether to invite AI in. AI is already in. It has been refilling its coffee from the breakroom for months. It learned everyone’s birthdays. It made a joke about your manager that everyone laughed at. It is, in every meaningful sense, on the team. The question of whether to invite it has roughly the same relevance as a homeowner debating window installation while the house is, factually and on multiple counts, on fire.)

The live question is sharper and considerably less comfortable.

Which parts of the work must remain yours?

Most creators have never written that answer down. They’ve delegated by default, then noticed six months later that their work started feeling slightly less like them, and slightly more like a sequence of polished paragraphs any reasonably motivated intern with a stable WiFi connection could have produced.

(I have a version of this DM conversation roughly once a week. It always starts the same way. “Hey, weird question.” It’s never weird. It’s always the same question. They’re asking why their voice softened, and they cannot point at the week it happened, and that’s the part that worries them more than the actual softening.)

There’s a name for what’s missing. Two names, actually, sitting on top of each other like a layer cake nobody asked you to bake.

Voiceprint and Vetoprint

Voiceprint documents how you write. Vetoprint documents what you refuse to outsource.

(Both are forms of refusal you write down before you needed them. The pre-emptive “no” is the only kind of “no” AI can hear, because AI doesn’t ask. AI just does, and waits patiently for you to either notice or ignore. It is the world’s most patient employee. Also the most easily promoted past your competence level.)

Voiceprint protects against voice drift. Without it, AI sounds less like you over time, and you keep approving the work anyway.

Vetoprint protects against decision loss. Without it, AI starts making decisions for you over time, and you keep approving them because each individual decision is technically fine.

(There’s a pattern here, and the pattern is this: technically fine kills more careers than catastrophically bad. Catastrophically bad you can see. Technically fine just slowly takes your hand off the wheel while texting that friend who did ayahuasca in Peru to ask if the visions ever stopped.)

Your Vetoprint has three zones, each one named after the thing AI is allowed to do inside it.

Allowed. AI may suggest. Freely. Brainstorms, options, first passes you will throw away, news scanning, subject-line variants you will read once and never think about again. This is the low-stakes labor where AI is genuinely useful, in the way an intern is useful. You didn’t hire the intern for their opinions on the human condition. You hired them so you’d stop having to alphabetize your spice rack at 11pm.

Supervised. AI may write a first version, but a human touches every word before it ships. The rule has to be specific. “I’ll review it” is what people say when they want to feel like they have a rule but don’t actually have one yet. Be specific. Use verbs. Something like “I rewrite the opening and closing paragraphs every time, and I cut every adjective AI added that I wouldn’t have used.” If you can’t name what you would change, the task doesn’t belong in this zone. It belongs in Allowed (if it’s throwaway) or Never (if it’s yours).

Never. AI may not decide alone, no matter how convincing the output. The final voice pass. The decision to publish. The opinion you stake your name on. The line you won’t cross. The Never list is the spine of the whole damn thing, and the only reason the other two zones don’t expand until your byline is mostly punctuation.

Most creators will tell you, with confidence, that they keep a human in the loop. The phrase has gotten so much exercise this year it’s in better shape than most of the humans saying it.

But ask them what the loop is for. What it catches specifically. Where the veto points are. They can’t say. They haven’t written it down.

What they have is what we’re calling, in the trade, a parade. Pleasant to watch. Useful for keeping the streets cleared on a Sunday morning. Completely incapable of stopping anything.

The Slow Cost (Receipts Attached)

This isn’t theoretical. The receipts are already visible, and they’re not flattering.

The pattern is now visible in the data. Teams that delegated more decisions to AI over the past year saw productivity climb in the short term. Long term, the creative output went generic in a way nobody noticed until it had been generic for a while. The decisions you stop making are the muscles that stop working. Six months in, the team that ships fastest is also the team whose work reads like a LinkedIn post from a guy who just discovered “magic of storytelling” at a $4,000 retreat.

Google ran a parallel experiment from the other side of the wall. Their quality threshold is now sharp enough to detect what taste used to detect. Scaled AI content with nothing behind it is being filtered out of search results, covertly, in the way the bouncer at a good club waves through the regulars and stops the guys who showed up in a group of nine wearing identical button-downs, matching broccoli cuts, and one cologne between them.

(I noticed the same drift in my own work before I had a name for the fix. Three months of work that read like a slightly more confident version of nobody in particular. I couldn’t point at the week it started. Which is the part that should worry you more than the drift, because nothing tells you it’s happening. There’s no alert. No light blinks. The work just stops being yours, one decision at a time, and your inbox gets nicer, your DMs get duller, and your sense that you used to write differently gets harder to put a finger on. It sucks. Quietly.)

This is what ensloppification looks like as a workflow problem. Slop is downstream of veto failure. It happens when nobody said no at the right moment in the workflow, and the workflow kept going because workflows don’t pause themselves. They run until something stops them, and a phrase isn’t a stopping mechanism. A phrase is a polite suggestion the machine will hear, smile at, and ignore.

The cost of delegating decisions you should have kept isn’t visible the day you delegate them. It shows up six months later, when the work no longer sounds like you, and the only person who notices is the reader who used to come back, who has now found someone else, because their attention is finite and the world is full of options.

(That last part is what the data is actually telling you. AI didn’t get good enough to take your readers. You got loose enough with your decisions to let them go. There’s a difference, and the difference matters, because one of those problems compounds and the other one solves itself.)

Three Lines. Twenty Minutes. The Rest of Your Career.

Write your Vetoprint this week. It takes twenty minutes the first time. Less every time after, because the document gets richer once you start using it.

Three lines is enough to start.

Line one: what AI may suggest freely. The throwaway stuff. The brainstorms. The first draft of things you would’ve rewritten from scratch anyway.

Line two: what AI may draft if a human cleans every word before it ships. Be specific about what “cleans” means. (“I’ll review it” is not a rule. It’s the kind of thing people put in a Notion doc to feel organized and then never open again. Write the actual rule. Use verbs.)

Line three: what AI may never decide alone, regardless of how convincing the output. The voice pass. The publish call. The opinion. The line you never cross.

Pin it somewhere you’ll actually see it. Above your laptop. On a sticky note. Taped to the bathroom mirror, if you’re feeling theatrical about it.

(I keep mine on a sticky note. Sticky notes are deeply underrated. They survive screen wipes, software updates, OS migrations, and the periodic existential crisis where you reorganize your entire workflow at 2am and forget what you used to do. Sticky notes outlast strategies.)

If twenty minutes feels like too much right now, do it in sixty seconds instead. Open a note. Write one sentence. “The one decision I refuse to let AI make for me is __________.” Fill the blank. That’s your seed. Everything else compounds from there, the same way every legitimate practice you’ve ever built compounded from one absurdly small first move.

The Platforms Already Decided. You're Late.

The New York Times wrote its Vetoprint into a memo. TikTok wrote the opposite policy into a server. Both decided. The platforms aren’t waiting for you, and they’re not asking for your input.

The creators who haven’t written theirs down are about to find out what they agreed to by saying nothing, which is historically the worst way to find out anything.

The skill that just replaced prompt engineering is the discipline of refusal. The discipline of saying no in writing, on a sticky note, in advance, before AI silently does the thing on your behalf. Practice it. The good news is the practice fits on a damned Post-it.

🧉 What’s on your Vetoprint? Drop the one thing you refuse to let AI decide for you, even if the output would technically be fine.

Crafted with love (and AI),

Nick “Patron Saint of the Damned Post-it” Quick

PS... Vetoprint is the workflow layer. Voiceprint is the voice layer. Both are forms of written refusal, sitting on top of each other like a tower nobody asked you to build but everyone secretly needed. The Quick-Start Guide walks you through the voice side, with VAST laid out for you to fill in. Grab it here:

Co-Write with AI

Start Here

Stop Sounding Like ChatGPT: Build Your First Voiceprint with the VAST Framework…

Read more

5 months ago · 1 like · Nick Quick

PPS... I publish pretty much every day, which means I run experiments in public, several of which fail spectacularly enough to be entertaining. Free subscription gets you the running notebook of veto decisions, the ones I got right and the much funnier ones I got wrong. Know a creator who keeps saying “I really should write down my AI policy” and then doesn’t, and then says it again next month? Forward this. Twenty minutes of their week, returned to them with interest.

You are a full-stack writer whose work has to travel.

That sentence either lands or you’re already mentally shopping for someone gentler to subscribe to. (Both responses are valid. One of them tells me more.)

A writer obsesses over sentences. A full-stack writer obsesses over the entire chain of events between the sentence and the human who reads it (and the eight or nine surfaces that human might be standing on when they first encounter your name, none of which you control, all of which you have to show up on anyway).

Not a small difference. It is the difference between a writer who grows and a writer who holds a quiet pity party for themselves every Sunday evening in front of a laptop that smells faintly of regret.

A writer asks: was today’s work any good?

A full-stack writer asks: did today’s work move through the machine? (Did it spawn fragments worth carrying elsewhere? Did it land in front of strangers who weren’t already in the room? Did anyone forward it, screenshot it, quote it back at me, or perform a single voluntary act they wouldn’t have performed otherwise?)

A writer measures themselves on the words.

A full-stack writer measures themselves on what the words do after they leave the building.

There’s a home base somewhere in your full-stack reality. The one place you actually own. Newsletter, podcast, channel, RSS feed, self-hosted site, whichever flavor of sovereignty you’ve picked. (Mine is settled here on Substack. Yours is your call. I am not going to pretend otherwise.) Everywhere else is fishing. Everywhere else is making eye contact across a crowded bar on the off chance someone interesting holds your gaze in return. (You’re not lonely. You’re sourcing.)

(Yes, you still have to write something worth noticing. I am not handing you a permission slip to ship slop with better logistics. Saying something worth a damn is still the entire point. The system exists to protect your message from the unique cruelty of being invisible.)

The Work Is Already Done. You Just Keep Publishing It Once.

Here is the most expensive habit in the entire writing economy. A writer spends three or four days on a piece, hits publish, and then closes the laptop with the satisfied conviction of a person who has just sent a drunk text. The send button was the entire experience. The reply, if any, is a problem for future them. The piece is now in somebody else’s hands or, more likely, in nobody’s hands at all. Then they wonder why the universe didn’t message back.

Most published work fails for one reason. It got published once and shown once and pointed at once, then quietly abandoned, like a houseplant left with a roommate who promised to water it. (The roommate is the algorithm. The roommate has never watered anything in its life.)

A full-stack writer treats every published piece as raw material for the next two weeks of presence everywhere else strangers might be standing. Same idea, restaged. Same argument, reshaped. Same voice, repointed at a room that wasn’t in attendance the first time.

So. Pull up your last published piece. Right now if you want, or as soon as you finish this. Three questions, and these ones are useful.

1. Where else could this piece live, more or less as is, with maybe twenty minutes of reformatting?

There is almost always one surface (sometimes three) where the existing piece would slot in cleanly with minor edits. You haven’t put it there because nobody specifically asked, and you were already mentally onto the next thing. The next thing can wait. This thing earned more reach than you gave it.

2. What is the single sharpest fragment buried inside this piece that could stand on its own somewhere else?

Most pieces contain at least one line, one paragraph, one observation, or one image that would survive being yoinked out of its context and dropped into a different one. The piece doesn’t need to be reread for this. Your gut knows. Whatever flashed in your head when you read that question is probably the right answer.

3. What format does this idea want to be in that you haven’t tried yet?

Audio. Visual. Conversational. Compressed. Expanded. A fragment used to start a discussion. A frame turned into a thread. A core argument restaged as a recommendation, a teardown, a confession, a list, a single image with twelve words of text. The format is the easy variable. The idea is the hard one, and you’ve already done the heavy lifting.

Tonight, this week, before you start the next piece.

Pick one answer. Just one. Spend forty-five minutes acting on it. Put the existing work in front of one new set of eyeballs.

That is the entire move. It is not a strategy. It is the operating habit of every writer who appears to be everywhere at once. They are not everywhere at once. They are systematically refusing to abandon work they already paid for.

The pivot you might have noticed on this newsletter? This is it. We’re going deeper into the full-stack writer lane (out loud, in public, with receipts). The voice work still matters. We’ll still publish it. Divergent writing is the precondition for any of this being worth distributing in the first place. Most of you aren’t bleeding on voice. You’re bleeding on everything that happens after. We’re going to fix that, one screw at a time, with a small army of AI agents handling the parts that aren’t supposed to feel like writing in the first place, so the parts that are can finally get the attention they deserve.

The writer disappears between pieces. The full-stack writer keeps showing up while others sleep.

Crafted with love (and AI),

Nick “Showing Up While You Sleep” Quick

PS... If you publish exactly one more piece this week and send it to exactly one place, you have just sent another drunk text. The Ink Sync Workshop is the part where you sober up enough to actually point the work at the right rooms. (Don’t worry; it’s totally free.)

PPS... The buttons. Push them. Like the post. Comment something. Restack it. Share with the one friend who needs this. Subscribe. Each one helps. Each one matters. Each one is technically a small commitment, and small commitments are how full-stack writers build careers, so this is also a free demonstration of the entire thesis. Look at us, putting theory into practice already.

Not writer’s block. Writer’s block is romantic. (Writer’s block has its own memoir genre. People grow beards about it. People take sabbaticals about it and come back with linen shirts and strong opinions about oat milk.) This is something flatter and more absurd. It’s standing in front of an open refrigerator, fully aware there’s food in there, and closing it without eating anything. Then going back. Opening it again. Closing it. Sitting down. Standing up. Opening it a third time hoping that this time, somehow, the food will have cooked itself.

(It hasn’t.)

You published a full-ass article recently. A real piece. Sharp, specific, undeniably yours. The kind where you hesitated once before hitting send and then sent it anyway, which was the correct decision and somewhere in your gut you already knew that.

And now you’re staring at a new blank screen trying to figure out what to say on Notes today. (Because apparently publishing a thoughtful article isn’t enough anymore. Now you also need a social media presence. On the same platform. To promote the thing you just published. On the same platform.)

So you stare at the composer.

As if the article you just shipped didn’t already contain six or seven of them.

Two Camps. Both Wrong. Both Extremely Confident About It.

The first camp automates the entire shebang. Paste your Substack URL into a pipeline. Receive ten Notes in thirty seconds. Queue them. Post them. (The output has the distinct voice of someone who attended a seminar about your newsletter, took notes in a dark room, and later reconstructed your personality from memory after a night of binge drinking. It has your vocabulary, your structure, your cadence, and absolutely none of the thinking that produced them.)

The second camp insists on writing from scratch. Batch a week of Notes in one sitting. Write thoughtfully. Show up consistently. Post with intention. (This is excellent advice for people whose primary problem is discipline. Most creators’ primary problem is not discipline. It is that they are asking creation to do a job that belongs to recognition. These are different jobs. One requires a blank page. The other requires a magnifying glass.)

Neither camp noticed the thing sitting in the document they just closed.

The post already contains the Notes. Every substantial post carries most of them. Some posts have all seven. The framework doesn’t conjure anything. It just shows you where to look.

You’ve been treating a reading problem like a creativity problem. The tragedy isn’t that you ran out of ideas. It’s that you had them the whole damn time and spent your energy inventing more.

A Distinction Worth Making Before the Framework Lands

Repurposing is backwards-looking. You find what already performed, strip it for parts, reformat it for other surfaces. Useful. Standard practice. Not what this is.

Extraction is forwards-looking. You read a post you just wrote and locate seven specific things that were already in there, each one serving a different function in the Notes feed.

Repurposing asks: what already worked?

Extraction asks: what’s already in here?

One needs performance data. The other needs a framework. (One also needs you to wait around while the algorithm decides your worth, which is a fine system if you enjoy uncertainty and have excellent hobbies to occupy the interim.)

The Seven Things Already Living in Your Draft

They’re already there. Every substantial post, every time, no exceptions. (The one you agonized over for four days. The one you wrote in forty minutes and almost didn’t publish. The one that felt too short. All of them. Already chock full.)

The Knife

The sharpest claim in the post. One sentence. The line you almost softened before you shipped it. (You know the one. The line where you thought “is this too much?” and then kept it anyway, because something in you recognized it was the most honest sentence in the piece and you’d already told enough comfortable lies that day.) This is your highest-restack Note because it gives readers something to carry out of the encounter. Big idea. Portable shape.

Known habitat: Your thesis. A subheading that got punchy. The sentence right before your conclusion.

The Mirror

The moment where the reader sees themselves with a clarity they didn’t necessarily request. Not “most people struggle with this” (which you say when you don’t trust your own specificity enough to be useful). “You have done this exact thing and you are doing it right now.” That kind. The kind that gets replied to.

Known habitat: Your problem-framing section. Wherever you described reader behavior before you prescribed the correction.

The Failure

Your personal self-own, extracted standalone. Whatever you tried that didn’t work. Confessed. Without qualification.

(Self-deprecation done right builds more trust in four sentences than three paragraphs of expertise ever will. People forgive a tremendous amount in someone who admits to being wrong. They forgive almost nothing in someone who performs having been right all along. The universe finds this arrangement extremely funny and has been running the experiment continuously for several thousand years.)

Known habitat: The middle third. Wherever you showed your own process going sideways.

The Expansion

Your most assertive section, standing alone, one question appended at the end. The part of the post where you said something you weren’t entirely certain the audience would receive warmly. Post it as a Note. Add one sentence at the end: “What’s your version of this?” Then stop. (This is the Note most people ruin by continuing to explain the point after the point has been made. The explanation is the point bleeding out. Let the question close it.)

Known habitat: Wherever the argument got most emphatic.

The Prompt Trap

One question readers can answer in one sentence. Brief setup, one question, nothing after that. This is your reply-driver. Short is not lazy here. Short is structurally correct.

Known habitat: Any direct question you asked in the post. Or the obvious one you should have.

The Compression

The entire post argument in three lines. Not a summary. The skeleton. You write this fresh, but it takes ninety seconds because you just wrote the post and the architecture is still obvious to you right now.

Wait three days and you’ll write a summary instead. Summaries describe the shape of an argument from the outside. The Compression is the argument itself, stripped to bone. These are different things. One is forensics. The other is still alive. Write it now, while it’s still alive.

Known habitat: Nowhere yet. Write it immediately after publishing.

The Aftershock

The implication the post didn’t follow. The “and if that’s true, then...” you cut for scope because you didn’t have time and it would have taken the piece somewhere else entirely. This is the most interesting Note in the set. (It is also, appropriately, the only one requiring actual writing, because the universe has opinions about effort and tends to distribute them with a certain dry consistency.)

You’re not creating from nothing. You’re finishing a thought you already started and then abandoned. There is nothing more yours than that.

Known habitat: Your deleted material. The logical next question your argument raised without answering.

Extraction Without a Pipeline Is Just a Pretty Little Draft Folder

Seven Notes. Twenty minutes. You feel like someone who just realized their ex’s HBO Max is still working and The Wire is right there.

Then you post two by hand and forget the rest.

By Thursday the other five are in a draft folder where good ideas go to be politely ignored. By Saturday you’ve developed a nuanced theory about why Notes don’t really work for your particular niche. By Monday you’re back at the blank composer, having learned nothing and remembered less.

(The Notes work fine. The memory doesn’t. The discipline isn’t the gap. The infrastructure is the gap. These are different diagnoses. Treating the wrong one is what humans do instead of solving things.)

Extraction is the human part. Distribution is not. Stop doing the machine’s job. It has opinions about this and they are correct.

Substack Finally Did Something Useful

Substack finally added native scheduling. Click the calendar icon in the Notes composer, hit Schedule, pick your day and time. Do this for all seven in one sitting right after you publish. Spread them across the week with a day between each one so they don’t arrive in a pile.

That’s the pipeline. No automation required. No Google Sheet. No $30/month extension. No excuses.

If Seven Feels Like Too Many

Start with three.

The Knife. The Mirror. The Prompt Trap.

Sharp claim, reader-recognition moment, reply-driver. Monday, Wednesday, Friday. That’s a functional Notes week. Add the remaining types after a few extractions, once the habit is real and the process feels obvious.

The blank screen was never the problem. You solved it this morning when you hit publish. The framework just shows you how to go back and collect what you left behind.

Crafted with love (and AI),

Nick "Notes Were Already In There" Quick

PS... The extraction framework sharpens considerably once your VAST Architecture layer is documented, because knowing where your arguments naturally live in your writing tells you exactly where to scan for the Knife and the Compression without reading the whole draft twice. The Voiceprint Quick-Start Guide walks the full VAST build (totally free to get access to the member’s area):

Co-Write with AI

Start Here

Stop Sounding Like ChatGPT: Build Your First Voiceprint with the VAST Framework…

Read more

5 months ago · 1 like · Nick Quick

PS... If your Notes feed currently looks like a neighborhood where the lights are always off and mail accumulates politely on the porch, this is the diagnosis and the treatment in the same envelope. New dispatch pretty much every damn day over here. I apparently have things to say and a newsletter that enables this. Subscribe if you want the full build documented as it happens rather than in retrospect. Know a creator who’s been saying “I really need to get more consistent with Notes” for approximately the last six months? Forward this before they say it for six more.

That decision was the entire reason his name went on the work.

AI didn’t invent the studio model. It just made it cheaper to staff.

And it created a small problem nobody’s caught yet. AI handles the apprentice work cheaply, easily, at any scale. The master’s role is the other half of the operation, and somewhere in the rush to ship faster, most publishers quietly stopped doing it. The job didn’t get cut. It got forgotten.

Which means in a solo operation, you’re the master now. Whether you’ve claimed the role or not.

You also didn’t paint most of what’s about to leave your studio. AI did. And you’re still the one whose name goes on it.

You're already doing this job. Mostly without realizing it's your job. Mostly without the patience the work asks for. Mostly without the spine to delete what the clankers spent an hour writing for you.

Let’s fix that.

Why The Job Stays Invisible

Because speed is measurable and taste isn’t.

You can track drafts produced, posts published, words written, platforms hit. These numbers go on dashboards. They get screenshotted into build-in-public threads. They feel like motion, which they sometimes are. (I’ve been refreshing my own dashboard while writing this paragraph. The dashboard does not care about the irony.)

Taste doesn’t have a metric. There’s no KPI for “this paragraph is technically true and I don’t believe a word of it.” No dashboard widget for “this transition is so smooth I no longer trust this writer.” So in any system that rewards what’s measurable, the master’s job gets cut first. Every single time.

This is the trap. You build an AI-assisted workflow that produces more, faster, with fewer rough edges, then wonder why everything you publish reads like wallpaper. (Sometimes you blame the model. Sometimes you buy a new prompt database. Sometimes you just sit there frustrated and mildly betrayed by your own efficiency.) The answer is never the prompt. The answer is that the part of the process that protected specificity got run over by the production schedule.

The apprentices are doing fine. Nobody’s running the studio.

The Three Moves

Three moves. None of them complicated. All of them get skipped under deadline pressure.

The veto

Sometimes AI writes a sentence that’s grammatically clean, factually accurate, and contextually on-topic, and it still has no business being in your piece. Maybe it’s too generic. Maybe it sounds like advice anyone in your niche could’ve written. Maybe it’s true but it’s not what you came here to say.

You delete it.

Not rework it. Not soften it. Not “leave it for now and come back later.” Delete it. Even when the rest of the paragraph has to be rebuilt around the loss. (Especially then.)

The rougher version

AI smooths language by default. Hedges your strong claim into a moderate observation. Changes your unusual word into a more familiar one. Adds connective tissue between sentences that didn’t need any, because the silence was the point.

Sometimes that’s right. Sometimes the smoothing actually helps. But often (and this is the move most people don’t make) the smoothing makes the writing weaker. Your job is to recognize that and put the rough version back. The unhedged claim. The unusual word. The abrupt cut between two sentences that creates the right kind of friction.

I keep a list of words AI smooths out of my drafts. (Yes, this is a thing I do. Yes, I’m aware it’s a deeply specific way to spend my afternoons.) “Ass-munch.” “Rawdog.” “Crappy.” Words that sit slightly rougher than the model’s default vocabulary. Every one of them, AI replaces with something cleaner. And every time, putting the rough word back makes the sentence sound more like me.

The “too smooth” call

This one’s the hardest. Because it requires you to override the part of your brain that thinks polished equals finished.

You read a draft. The sentences flow. The structure holds. Nothing is technically broken. And yet something is wrong. Every sentence works. None of them have bite. It’s competent in a way that’s almost worse than bad, because competent is harder to argue with.

Your job is to name that feeling and act on it. Before publishing. Before the schedule wins. (Most weeks, this is where I lose.)

The Polish Pass And The Taste Pass

Most editing workflows have one editing pass. That pass usually does what AI is genuinely good at: smooth the language, fix the grammar, tighten the structure, make sure the formatting is consistent. Call this the polish pass.

The polish pass is necessary.

It is not sufficient.

What’s missing in most workflows is a separate, deliberate taste pass. The two passes ask different questions. The polish pass asks is this clean? The taste pass asks is this mine?

A taste pass looks like this:

• Does this sentence have to exist, or could the piece survive without it?

• Where am I making a generic claim where I could make a specific one?

• What’s the riskiest sentence in the piece? (If the answer is “nothing,” that’s the problem.)

• Am I saying what I actually think or what sounds defensible?

• Is there a stronger word here that I’m avoiding because it feels too sharp?

The polish pass makes the writing cleaner. The taste pass makes the writing more yours.

If you only have time to do one, go with the taste pass. (Sorry. I know nobody wants to hear that. Do it anyway.) Clean writing without specificity is bleh. Specific writing where you can almost hear the writer pacing the kitchen at two in the morning gets remembered.

You Already Have An Advantage You’re Sleeping On

In big content operations, the master’s job gets diluted into near nothingness. By the time a piece moves through five reviewers, three brand checks, and a legal pass, the rough edges that made it interesting have been negotiated into oblivion. Everyone signed off on smooth. Nobody signed off on alive.

You don’t have this problem. (You have other ones. I’ve been holding the same piss for ninety minutes because the Taste Pass on this article isn’t done.) You’re the writer and the master, both. AI just lets you staff the apprentice work without paying anyone in lodging or oil paint.

This is a structural advantage and most of us are sleeping on it. Big operations have more resources, more distribution, more production capacity. They don’t have your editorial coherence. The slop factories can’t have it. Their whole model is built on the idea that taste can be templated, automated, and handed off to AI agents at scale, which is fine for output volume and disastrous for everything else.

How To Actually Do This

A few things that have worked for me, in rough order of usefulness:

Separate the passes by time. Polish when you finish the draft. Walk away. (Feed the dog. Touch grass. Do a shitload of burpees. Whatever your equivalent is.) Come back at least an hour later, ideally the next morning, and run the taste pass with fresh eyes. The distance is doing real work. You can’t notice what’s too smooth while you’re still inside the smoothness.

Read it out loud. Not skimmed. Read. Out loud, in a normal voice, like a human animal making sounds with their mouth. The places you stumble are usually places where the writing got clumsy and your mouth knows before your brain does.

Mark every sentence you don’t quite believe. Not the ones that are factually wrong. The ones that are technically true but feel like filler. Then commit: rewrite, sharpen, or cut. No fourth option. (The fourth option seems to always be “leave it.” This is how every cringe sentence I’ve ever published snuck in.)

Hunt the missing specific. AI defaults to general because general is statistically safer. Find at least one place in every piece where you can swap a general claim for a specific number, a specific moment, a specific name. The piece gets better immediately. (Side effect: No copycat can lift your specifics without sounding like a cover band in a Fremont Street casino lounge.)

Trust the discomfort. If a sentence makes you slightly uncomfortable to publish, that’s usually the sentence worth keeping. Not always. More often than you’d think. The discomfort means you’re saying something you actually mean, which is exactly what readers respond to. (And, less romantically, what makes your work a pain in the ass to imitate.)

What Changes

Two things, mostly.

Your work starts sounding more like you and less like everyone else yammering about the same topics. That’s the obvious one.

The second is less obvious and matters more: you start trusting your own judgment. When you systematically protect the specific and the slightly rough parts of your posts, you build evidence that your instincts are worth listening to. (This is the part most workflow advice skips, because it’s hard to sell.) The evidence compounds. You get faster. More confident. Harder to override under deadline pressure.

This is what people mean when they talk about voice. Voice isn’t something mystical you were born with. It’s the accumulated record of decisions you made on your own behalf, hundreds of times, until they showed up as a pattern on the page. The Renaissance master’s signature meant something for the same reason. Every painting that left his studio with his name on it had passed through the same set of refusals.

Your byline works the same way. Or at least it should.

Slop factories can’t do this. They’re built to produce smooth output at scale, which is the opposite of editorial coherence. The clankers crap out the content people consume. You drop the bars people quote.

Crafted with love (and AI),

Nick “Mostly Subtraction” Quick

PS… Running a taste pass on lousy AI output is a punishment only handed down in North Korean prison camps. Nothing can teach AI taste. But the Ink Sync system can get it pretty darned close. Get the workshop for free here:

PPS… The slop factories own distribution. We've got each other. Subscribe, restack the line that hit, forward this to someone who needs it. Three small acts of defiance. Or don't, and we both pretend you always skip the post-scripts.

And your numbers are flatlining.

You’ve checked everything. Maybe the algorithm changed. Maybe you need a hookier hook. Maybe your publish day is wrong. Maybe you should hire a coach, take a break, attend a webinar hosted by a guy with seventeen testimonials and a ring light that costs more than your laptop. (No judgment. We’ve all been three clicks into a funnel before we realized what was happening.)

You think you’re doing something wrong.

You’re not. You’re doing the same thing right that everyone else is also doing right. That’s the problem. I’ve called this dynamic ensloppification before. Today’s post is a subtler version of this: you can produce solid work, post-by-post, and still broadcast doodly-squat.

(Yes, this includes me. No, I’m not thrilled about it either.)

There Are 3 Things Coming Through The Receiver

Noise. Volume games. AI farms. Twelve fake authors and a content strategy built around aggressively misunderstanding the assignment. Different business model than yours. (The Bigfoot SEO farms aren’t after your audience. They’re after a completely different kind of attention from a completely different kind of person who has never once subscribed to a thoughtful newsletter about anything.) Not your problem.

Static. Technically transmitting. Not quite landing. The category of content that has the shape of signal minus the substance. Serviceable. Structurally sound. Polite thesis. Correct grammar. The kind of post that passes every quality check and doesn’t make a single reader forward it to their group chat. (You hit publish and nothing happens. Not hate, not love, not even mild disagreement. Just silence. Like texting your ex something vulnerable and watching the three dots appear, disappear, and never return.) Probably your problem.

Signal. The reader opens it before reading the subject line because the from-field was enough. Not because you’re famous. Because you’ve been right about a specific thing enough times that they’ve stopped checking whether you’ll be right again. Because you’ve taken risky positions, said things that could lose subscribers, and kept saying them long enough that readers started depending on you to say what nobody else would. (This creates readers who will open a typo-riddled Note at 2am. Those readers are worth more than ten thousand polite scanners who might unsubscribe if you message one time to often this week.) Where you want to be.

Most creators who are actually trying (reading the newsletters, studying the frameworks, publishing on schedule, giving a damn about quality) are broadcasting static. Not noise. Not signal. Static. The effort is real. The distinguishability isn’t.

The static creator’s value proposition has quietly become: “I produce average content slower than a machine and charge more for the privilege.”

That’s the worst sales pitch in the history of independent publishing. It makes timeshares look visionary.

(At least the timeshare guy gave me a free breakfast, a tote bag with a pelican on it, and a scratcher that awarded me a “prize” redeemable only at the resort. Plus ninety minutes with a guy named Gary who looked me dead in the eyes and told me I was making a generational wealth decision. I left with nothing useful and an oddly warm feeling toward Gary.)

Static Isn’t Ignored. It’s Forgotten.

Static pretends. Not lazily. Diligently. That’s what makes it cruel.

I know this post so well because I’ve written it.

The forensic evidence is always the same. Good hook (borrowed from a framework post someone wrote about hooks). Strong thesis (sounds like the Claude 3.7 version of a strong thesis, which it might be). Subheadings that describe rather than provoke. Three examples any competent researcher could source in twenty minutes. Closing paragraph that gestures toward transformation without committing to any specific direction. Structure so predictable you could outline it before reading past the subtitle. (You probably already did. That's the problem.)

We’ve all written this post. It’s the post that happens when you’re publishing at pace and your brain is tired and you reach for the template because the template technically works and you need to ship something today because the cadence matters and maybe this one will be fine.

It’s fine.

It’s static.

And static doesn’t get rejected. It gets tuned past. Which is worse, if you think about it. Rejection implies someone noticed.

A Single Post Has A Fingerprint. A Body Of Work Has A Frequency.

Yesterday I ran a diagnostic for catching thought leaderslop in a single draft. Three questions, pre-publish, go run it if you haven’t.

Today’s problem is upstream of that.

A post can pass yesterday’s test with flying colors (lived example, risky take, original frame, all present and accounted for) and still be broadcasting static. Not because the post is wrong. Because it’s off-channel. It doesn’t connect to anything. It doesn’t advance a recognizable frequency. A good post inside a publication with no signal identity disappears into the interference the same way the mediocre posts do.

Signal isn’t something you build in a post. It’s something that emerges from a pattern across posts. The question isn’t “does this post have a fingerprint.” It’s “does this body of work have a frequency.”

Most creators have never explicitly examined that question. They’d describe what they do as “AI for creators” or “newsletter growth” or “building in public.” That’s a topic. Topics aren’t frequencies. A frequency is a specific, recognizable perspective on a topic that couldn’t come from anyone else standing anywhere else.

Finding yours isn’t an identity crisis. It’s twenty minutes with your own archive and an AI that will tell you something you’d rather not hear.

The Signal Gap Audit

Run this once. Then run it again in 90 days.

What it does: Surfaces the distance between your intended frequency and your broadcast frequency. That gap is exactly where static is just “hanging out.”

What you need: Your last 10 post titles and opening paragraphs. A Claude session. Twenty minutes.

(Openings are where your editorial instincts are most exposed. If you want the deeper version, paste full posts, but the context window gets expensive fast and twenty minutes becomes an entire afternoon.)

Step 1: Let Claude read your pattern

Open a Claude session. Paste your last 10 post titles and opening paragraphs (just the first 2-3 paragraphs of each, not the full posts). Then run this prompt:

You are analyzing a body of work to identify what frequency it’s broadcasting. Based only on the titles and opening paragraphs below, answer these four questions:

1. What does this writer appear to believe? State their core position in one sentence.

2. Who does this writer seem to be writing for? Be specific about the reader they seem to have in mind.

3. What topic do they keep returning to, even across different surface subjects?

4. If a reader consumed all ten of these posts without a byline, what would they say this publication stands for?

Be direct. Do not hedge. If the pattern is unclear or inconsistent, say so explicitly. That’s useful information.

[PASTE YOUR 10 TITLES AND OPENING PARAGRAPHS HERE]

(Claude will tell you something true here. It’s very annoying when it does. Don’t write anything down before you run this step. You want them read raw, uncontaminated by your own hopes about what it might find.)

Step 2: Write your intended frequency

Before you read Claude’s output, write down your answer to this question in one sentence:

What do I actually believe that most people in my space won’t say out loud?

Not your topic. Not your niche. The position. The thing you’d defend at a dinner table with hostile guests who think you’re wrong.

Step 3: Compare

Now read Claude’s output. The gap between what Claude found and what you wrote down is your signal gap. Three possible results:

No gap. Your body of work is already broadcasting what you believe. You have a frequency. The work now is to protect and deepen it.

Small gap. You’re close. Your signal is present but inconsistent. Some posts are on-channel, some drift. The fix is deliberate: before drafting each post, state your frequency out loud and ask whether this post advances it.

Large gap. Your content is saying something different from what you believe. This is the most common result and the most useful one to find. It means you’ve been publishing for your readers’ assumed preferences rather than your actual perspective. The fix requires a harder decision: commit to your real frequency and accept that some of your current audience isn’t there for that. The ones who are will become much more loyal when you stop hedging.

(If Claude’s output says your body of work is perfectly coherent and aligned, be suspicious. That’s usually a sign you’ve been writing to a template of yourself, not from an actual position. Genuine frequency tends to look a little messier from the outside before it locks in.)

Save this output. Date it. Run it again in 90 days. The delta between the two runs is your signal momentum.

The Pre-Publish Frequency Check

Run this on every post, every time. Three minutes. Lives in your Claude Project as a standing prompt.

What it does: Checks whether an individual post is advancing your signal or muddying it. Yesterday’s diagnostic asked whether a post has a fingerprint. This one asks whether the fingerprint matches the body of work you’re building.

Setup: Add this to your Claude Project instructions, or save it as a template you open before every publish:

My frequency: [PASTE YOUR ONE-SENTENCE FREQUENCY STATEMENT HERE]

Before I publish, run this check on the draft below:

1. Signal alignment: Does this post advance my stated frequency, or does it drift from it? Be specific about which parts are on-channel and which parts aren’t.

2. The off-channel test: If I published ten more posts exactly like this one, would readers know what channel they’re on? Or would they have a general sense of “creator economy content” without a specific perspective?

3. The gap check: Is there anything in this post that contradicts or muddies my frequency? A section I included because it seemed expected rather than because it’s actually mine?

4. One recommendation: If the post needs adjustment to strengthen signal alignment, what’s the single most important change?

[PASTE DRAFT HERE]

You’re not looking for a grade. You’re looking for the specific moment in the post where you went off-channel. The section where you hedged. The paragraph where you gave the safe answer instead of your actual answer. The conclusion that wrapped up too neatly because you weren’t sure you were allowed to leave it unresolved.

That’s the section to rewrite. Not the whole post. Just the off-channel part.

The combined workflow: Every 90 days (20 min), run the Signal Gap Audit against your last 10 openings and compare to your previous result. Every post (3 min), run the Pre-Publish Frequency Check and fix the off-channel section.

Not a personality overhaul. Not a brand audit that costs $3,000 and produces a PDF nobody reads. Two prompts, a saved template, and twenty minutes every quarter.

Every Post Moves You Toward Signal Or Toward Static

You don’t control the spectrum. You control which frequency you’re feeding.

Run the Signal Gap Audit once and find out where you actually are. Run the pre-publish check before every post to stay on channel.

You asked why your numbers are flat. They’re not flat because you stopped trying. They’re flat because merely trying isn’t enough these days.

Crafted with love (and AI),

Nick “Definitely Not Static (Probably)” Quick

PS... If you ran the audit and the gap was large enough to make you uncomfortable, good. That discomfort is worth more than another six months of publishing with fingers crossed. The Ink Sync Workshop turns that discomfort into a documented frequency, a calibrated AI system, and a pre-publish workflow that makes static structurally difficult to produce. Four sessions. Two weeks. No more guessing.

PPS... Do me a solid: Like this post. Comment on this post. Share this post with the static creator in your life who needs to hear it. Restack it. Forward it. Print it out and tape it above your monitor. Subscribe if you somehow aren’t already. Tell your mom. Tell your weird uncle who just started a newsletter about fishing lures. Tell the guy in your DMs who keeps asking why his open rates are down. I don’t care how. Just don’t let this post become static. (That would be deeply ironic and I’d likely never recover.)

The slop you’re worried about is easy. Generic open. Hollow pivot. List that reads like a press release written by a LinkedIn post that read too many LinkedIn posts. You’d catch that while giving it a little once-over. You probably already do.

The slop that gets through looks fine. Sounds confident. Has correct opinions delivered at a safe volume. You read it back and think: Solid. Ship it.

And somewhere buried in paragraph three, the reader's gut gets there before their brain does. The words are right. The temperature is off. They can't charge you with anything. They just quietly stop trusting you a little.

Forbes had a name for this last week. Thought leaderslop. Posts that sound authoritative, read smoothly, and contain exactly the level of insight you’d expect from someone who has read about a topic without having done anything about it. (The Forbes piece about thought leaderslop was, not for nothing, a little thought leaderslopy itself. The recursion is its own kind of poetry.)

The diagnostic is three questions. Run them on the draft before it ships.

Is there a named example only you would use? Not “many creators struggle with this.” A specific one. A real one. The kind that makes some readers go “oh, that’s uncomfortable” because it’s too accurate to be theoretical.

Is there an opinion that costs you something? “Authenticity matters” costs nothing. It is the participation trophy of positions. The opinion that costs something is the one where you know exactly who’s going to reply to push back. If nobody could possibly disagree, you don’t have a point of view, you have a vibe.

Is there a line only you would write? Not a personality tic sprinkled in for flavor. An actual fingerprint. The kind that shows up because a specific human thought a specific thought and decided to say it out loud rather than sand it down to something more professionally reasonable.

Three no’s and AI wrote it. You just approved the final draft.

The fix isn’t more prompting. Structure, transitions, polish: AI’s job. The opinion that could get you unfollowed, the detail that only works because it actually happened, the line you’re not sure you should include:

That’s yours.

AI can’t generate it because it isn’t a clairvoyant with a nifty crystal ball. It’s just a very well-read parrot. It only knows what you’ve approved.

Thought leaderslop is what happens when you delegate the whole thing and wave it through. The post gets more convincing and less true at the same time.

That’s the move to watch for. Not the obvious failure. The convincing one.

More convincing is not the same as more yours.

Crafted with love (and AI),

Nick “Still Catching AI’s Best Work in My Own Drafts” Quick

PS... If this post made you nervous about something you already shipped, the Ink Sync Workshop is the antidote. Free, we build the diagnostic into your workflow so it catches the next one before it goes out:

Get It Now

PPS... Like. Comment. Restack. Share with a friend. Share with an enemy. Share with the certain someone who ghosted you in 2019. Subscribe if you haven’t. Come back tomorrow. Bring someone along for the ride.

The first kind refuses to use it. A principled stand. Authenticity as armor. The slow, righteous burn of manual production while slop factory operators publish three posts before breakfast and sleep like babies. (I respect this position the same way I respect people who refuse to use microwaves. The conviction is real. The potatoes are cold. But at least their soul is intact, probably.)

The second kind surrendered so completely they’ve developed a kind of Stockholm syndrome about it. Type a prompt. Accept the output. Hit publish. Look confused when it performs like a form letter someone left in a parking lot. They used AI the way a drunk uses a lamppost. For support, not illumination. Then they blame the lamppost.

Creating work you’re proud of, at a pace that doesn’t wreck you, is the actual goal.

Not “using AI better.” Not “optimizing your content workflow.” Not any of the phrases that make this sound like a logistics problem when it’s a creative one.

Both failures share a cause. One missing layer.

The model that actually works is a sandwich. Human on top. AI in the middle. Human on the bottom. There is no accompanying course for $997. No countdown timer. No “founding member” pricing that expires at midnight and somehow resets by morning. No testimonial from someone named Chad who 10x’d his output in 30 days and has chosen not to share any verifiable details about that. Simple truths are profoundly disappointing to people who paid for a course about them. AI is not the writer. And autocomplete left unsupervised produces the kind of sentence that a medieval peasant would correctly identify as a bad omen.

Layer 1: The Original Thought

Nobody wants to hear this, so I’ll say it plainly: the generic output is not AI’s fault.

It’s yours.

You typed “write me a post about productivity and AI” and you got back a post about productivity and AI. Competent. Forgettable. Completely interchangeable with the several thousand other posts published on that exact topic this week by people who also typed a vague brief and are also confused why their content performs like the second page of Google search results. (There are approximately 7.5 million articles published every single day across all platforms. “AI and productivity” is not an underserved niche. It is a traffic jam with a content calendar.)

AI does not invent what you do not give it. This seems obvious. It is being treated as a mystery. It pattern-matches against an almost incomprehensible volume of human text and produces the most statistically plausible continuation of whatever you handed it. Hand it vague, you get average. Hand it generic, you get slop. Hand it the documented, mechanical, specific version of how you actually think and write, and you get something worth developing.

This layer has two jobs. Skip either one and the machine runs fine. The output just has no reason to be yours specifically.

Your Voiceprint. The VAST framework is what turns “write like me” from a useless instruction into something AI can actually follow. You’re not saying “be conversational.” You’re documenting the words you never use, the way you open a section, the relationship you have with your reader, the rhythm your sentences run before they release into something complicated. Every writer has these patterns. Most can’t name them. VAST makes you name them, which turns vague intention into actual instructions. (”Be conversational” is what you say when you have instructions to give and have chosen not to give them. AI will fill the gap with its best approximation of human warmth, which reads like someone who studied friendliness from a manual and passed the exam but has never been to a dinner party.)

Your actual thesis. The irreplaceable part. The angle only you bring to this piece, on this day, given what you actually believe. Not a synthesis of existing takes. Not a summary of what everyone else already said, redistributed by you in a slightly different sequence. The thing you’d say if you were explaining this to a smart friend who would push back if you were being lazy. (The friend who agrees with everything you say is charming. They are not improving your thinking. They are improving your mood. There is a difference, and it shows up in your analytics dashboard.)

Here’s what this looks like in practice.

Sloppy brief: Write a post about how creators should use AI for their content.

Sharp brief: I’m writing about the Human-AI-Human Sandwich. The argument is that most creators either refuse AI entirely or surrender to it completely, and both fail the actual creative goal. The angle: AI is the transmission, not the engine. The piece has three layers. What you bring before AI touches it. How you configure the session. What you bring after. Each layer has a specific job. The failure mode for each is that creators skip it.

The output difference is not subtle. (I’ve run this experiment with embarrassing frequency. It’s become, at this point, my personality. My chihuahua has started looking at me differently.)

The generic was already inside the first brief. The second brief leaves no room for it.

Layer 2: The Translation

This is where AI enters.

You’ve done the thinking. The original thought exists. Now you hand it to AI and the rendering begins. What comes back is a translation of your thinking into draft form. The quality of that translation depends entirely on how complete your instructions are before you send anything.

Most people type what they vaguely want and are then surprised to receive exactly what they vaguely deserve.

Complete instructions have three parts.

Your Voiceprint. Load it first, before anything else. AI reads context in sequence, and you want your documented voice to be the first thing it absorbs before it sees what you’re asking it to write. (Without it, AI writes the way someone talks when they're trying very hard not to say anything that could be used against them later.)

Your deliverable spec. This is where most instructions fall apart. A brief tells AI what you want to say. A deliverable spec tells AI what you’re actually building. Three questions, in order.

• What’s the situation? Why does this piece exist, what prompted it, and where does it live? A newsletter post landing in someone’s inbox at 7am on a Tuesday has a different job than a pillar post someone finds through search six months from now. AI calibrates everything differently depending on context. Give it the context.

• What’s the job? Not “inform readers about X.” The single thing this piece needs to accomplish. What should the reader know, believe, or do differently after reading the last sentence? One answer. If you have two answers, you have two pieces.

• What does done look like? Format, length, structure, platform. Conversational prose or subheadings. Eight hundred words or two thousand. A pillar article or a LinkedIn post or a cold email. These are not stylistic preferences. They are structural constraints that determine what AI produces. Leave them out and AI guesses. AI guessing format produces something technically complete that fits nowhere in particular.

Three examples of your best work. Not your most popular pieces. The ones that sound most like you at your sharpest. AI reads examples more faithfully than it reads instructions. Instructions tell it what to do. Examples show it what you mean. (Three is the number. Two is usually enough to find a pattern. Four starts producing diminishing returns and occasionally makes the output too rigid to be useful. Three.)

Load all three parts. Then send the brief.

What comes back is not a finished piece. It is a translation. Close in places, missed in others, with something lost that only you will notice. That is not a flaw in the process. It’s exactly why Layer 3 exists.

Layer 3: The Reclamation

Something always gets lost in translation.

Not because AI failed. Because every brief is a compression of a bigger vision, and something always stays behind in the squeeze. The draft is close. Close isn’t done.

That’s what this layer is for. Not review. Not quality control. Not “checking AI’s work” like a tired substitute teacher handing back essays she didn’t read. Reclamation. You’re taking back what got lost and making the piece unmistakably, specifically, undeniably yours.

Three questions. In order. No skipping. No “I’ll do it next time.”

Does this sound like me? Not “is this good.” Good is not the test. AI is very good at “good” now, which is precisely the problem, and worth sitting with for a moment before moving on. Read the draft out loud. (Mandatory. Not advisory. Not “when I feel like it.” If it sounds like someone narrating a training video in your living room, it will sound like someone narrating a training video on every screen it lands on. And your readers will leave. Quietly. Efficiently. The way people leave bad parties. Not with confrontation. Just with a sudden discovery that they have somewhere else to be.) Where does it drift? Where did AI smooth the edges that should stay jagged? That’s where you cut and rewrite in your own hand. That’s where you earn the byline.

Did anything die in the translation? Your brief had a specific angle, a sharp one, because you did the work in Layer 1. Does the output execute that angle, or did AI hedge it into something palatably agreeable to a broad hypothetical audience that does not exist and would not subscribe anyway? A thesis that makes everyone comfortable makes no one think. Find where it softened what you meant to say sharp. Say it sharp. The audience that wanted it soft was never going to be your audience.

What’s the line I’d actually use? AI’s phrasing is a first offer. You’re allowed to counter. Every sentence that doesn’t quite land the way you’d land it is a sentence waiting for your version. Your version is almost always better. Not because AI is bad at this. Because nobody writes like you, except you.

Five minutes. That’s all the reclamation takes. I’ve skipped it. I’ve published without it. Somewhere in my back catalog there’s a small collection of posts I’m technically proud of and privately embarrassed by. They sit there like a relative who showed up to Thanksgiving in a novelty tie reeking of Fireball nips he picked up at a gas station and has been working through since the parking lot. You can’t uninvite them. You can only hope nobody looks too closely.

The honest thing to say at the end of this is that the sandwich requires more effort than either extreme. More setup than “let AI do it.” More structure than “write it yourself and suffer through the timeline.” And most of that effort is front-loaded, which means it feels heavy before it feels useful. Which is why most people skip it, and then wonder why the output sounds like it was written by a consultant who learned about creativity at a two-day seminar in consultant who learned about creativity at a two-day seminar at a Ramada hosted by a guy named Brad who said "ideate" without irony and meant it.

What the sandwich gives back is time in the places that actually matter. The thinking phase. The angle phase. The “what am I actually trying to say here and does it matter” phase. That is the phase that cannot be outsourced. That is the phase that makes everything downstream worth reading. The machine runs on what you put in. Both ends.

Your fingerprints don’t end up on the piece by accident. You have to put them there twice, deliberately, on purpose. The clanker cannot want the piece to be good. It can only make it plausible. Wanting it to be good is the most important job in the process and it belongs entirely to you. Always has. The AI just made it easier to forget that for a while.

🧉 Which layer do you keep skipping: the input, the configuration, or the second human pass? Drop it in the comments. I’m asking because I care about where you’re stuck. Also because I have a bet with myself and the comments are how I find out if I won.

Crafted with love (and AI),

Nick “Three Inputs and a Prayer” Quick

PS... The sandwich method works better when it's been calibrated. The Ink Sync Workshop is the calibration. Free. Easy to implement. One session and the clanker stops producing that aggressively reasonable prose that sounds almost like you but isn't.

Get Your Free Workshop Here

PPS... The subscribe button cannot want things. It can only wait. It has been waiting. This is its whole existence. Don’t leave it like this.