I’ve always had a low tolerance for preventable problems. Things that don’t need to be difficult but somehow always are. Forgetting an umbrella. Leaving lights on. Realizing the heat’s been blasting all day while I’m out. None of it’s catastrophic, but it adds up to a background hum of inefficiency that quietly eats your time.

It started with wet socks.

Not metaphorically — literally. I was stepping out of a grocery store in Manhattan when the rain came down like a punishment. Everyone else popped umbrellas in one motion, but I didn’t have mine. I froze, holding two paper bags already starting to wilt. The wind turned sideways, the street overflowed, and within seconds my shoes made that miserable squelch that stays with you for hours.

If you’ve ever been caught in a New York downpour, you know the feeling. It’s not just being wet, it’s being defeated. You can’t sprint home without soaking your jeans, and no amount of scaffolding saves you. By the time I made it back to my apartment, I looked like someone who had fallen into the Hudson River fully clothed.

That night, as I blow-dried my shoes, I promised myself I’d never feel that stupid again.

I didn’t even know where to begin. I wasn’t a developer. I didn’t want to write code. I just wanted a digital version of that Post-it note you leave on the door — don’t forget your umbrella. Somewhere in a forum thread, I stumbled onto Make.com, a no-code automation tool that promised to connect different apps like puzzle pieces. I was cautious, but it felt approachable enough to try.

A few nights later, I had my first working prototype.

The logic was simple, at least in theory:

If the smart lock on my apartment door unlocks, then check the local weather.
If the chance of rain is over 20%, send me a push notification: “Grab your umbrella.”
If the temperature drops below 60°F, ping me again: “You might need a jacket.”

Getting it to work was a different story. I connected the wrong modules, broke the flow, retried. I remember sitting there at midnight watching the Make.com dashboard like a heartbeat monitor, hoping that when I unlocked the door, the system would finally respond. And when it did, my phone buzzed.

“Grab your umbrella.”

I laughed out loud. The code was janky, the data barely synced, but it worked. A tiny miracle. A problem, solved invisibly.

From there, I couldn’t stop.

I moved the setup into Home Assistant running on a Raspberry Pi, and started building more logic on top of it. With one press of the TV remote, the lights dim and the curtains draw themselves closed. Plugging my phone in at night becomes a “goodnight” command — lights off, thermostat down, silence. When my partner leaves the office, I get a small ping on my phone: time to start cooking dinner.

Friends come over and shake their heads. “You’ve lost it,” they say, laughing as the curtains move on their own. But then they stop and ask how it works. Because underneath the absurdity, there’s a kind of elegance. A home that responds, not performs.

I won’t lie: some of it got out of hand. I automated so much that I started automating things I didn’t even need. The system would flick the hallway lights on if it detected motion — even if that motion was my cats. I learned the hard way that automating without purpose doesn’t create freedom. It creates noise.

Not everything should be automatic.

That realization became its own quiet philosophy: find the real pain points, not just the possible ones. The goal isn’t to make everything effortless. It’s to make the right things disappear — the small, predictable frictions that steal focus and attention.

That shift from “make everything automatic” to “make only the right things automatic” became the foundation for how I build systems today.

Now, when I work with teams, I see the same pattern everywhere, their own versions of the umbrella problem. Small, constant frictions that make good work harder than it needs to be. Most teams don’t need new tools, they need invisible systems that clear a path for people to do what they’re already good at.

So we build them.

An employee checks a box, and a real-time Slack notification alerts the next person — no one waiting, no one guessing. When a deliverable goes out, execs who need to know are notified instantly — no chasing, no “just checking in.” Documents that used to take hours to create, sort, and label now generate automatically — clean, uniform, finished in seconds. And for the leaders who need to see the signal through the noise, we build daily briefings, surfacing exactly what matters that morning, and nothing else.

These aren’t magic tricks. They’re acts of respect.

Automation, done right, isn’t about replacing people. It’s about removing the parts of work that make smart people feel stupid.

When you build it right, it feels like the lights dimming as the movie starts — effortless, expected, right on time.

Not too long ago, I was standing on the 70th floor of One World Trade Center at a Make.com mixer. Floor-to-ceiling windows stretched across the skyline, the entire city laid out below in perfect grids of light. It was one of those rare moments where you can see the system you’ve been living inside — the movement, the structure, the potential — all at once.

I caught my reflection in the glass and thought about that soaked version of myself outside the grocery store two years earlier, the one trying to solve a $5 problem.

The same curiosity that wanted to stay dry now powers entire systems that keep teams from drowning in their own noise.

I didn’t plan that arc. I just followed the irritation until it turned into something useful.

Solve for one.
Solve what hurts.
That’s how scale actually begins.

I was watching The Studio on Apple TV the other night, a few hours after the news broke that Catherine O’Hara had passed away. The timing felt uncomfortably precise. I’ve been watching O’Hara my entire life. She spent it playing women who understood the performance even as they were trapped inside it — characters sharper than the worlds that needed them, often reduced to spectacle anyway. I’ll never forget how exact her comedy always felt. The timing, the control, the sense that every gesture was chosen. Even at her broadest, she made the work feel deliberate, as though the joke itself knew more than it was letting on.

The Studio operates in that same register. Recently renewed, it treats modern Hollywood as a system that keeps mistaking continuation for survival. It’s a love letter wrapped in satire, a peek behind the curtain of an industry that keeps trying to resuscitate itself through noise, leverage, and denial.

Bryan Cranston plays the studio head, a man who cares nothing for art and everything for outcome. He’s the embodiment of a machine that confuses activity with vitality — endlessly reassuring itself that profit is proof of life.

In the episode, the studio is hosting its big upfronts — that yearly ritual when distributors and theater owners gather to be sold on the new slate of movies.

It’s a circus, but a joyful one.
Movie stars crowd the stage, trailers play in perfect rhythm, speeches swell with just the right amount of hope.

The audience isn’t faking it; they’re having fun.
For a few minutes, everyone remembers why they came into this business in the first place.

And that’s when I hear it, tucked behind the raucous applause: a melody that pulls me straight out of the moment.

My ear perks up like a reflex. What is that?
The rhythm is impossible to mistake. Is it?
And then I laugh, quietly but audibly. Oh my God, it is.
Followed by the inevitable realization: Oh my God, this must be so expensive.

“There’s a Great Big Beautiful Tomorrow…”

They chose it because it fits.
The characters have just pulled off something improbable, and the song seals it — an anthem for survival disguised as a jingle about optimism. For a brief moment, everyone in that fictional theater believes again: that movies still matter, that audiences will still show up, that creativity isn’t dead yet.

The Carousel of Progress

Walt Disney’s 1964 love letter to automation — the ride he designed for the World’s Fair, later moved to Disneyland.

A rotating theater, one family living through each new age of invention: the electric age, the jet age, the imagined home of the year 2000. The family never changes; they simply move forward through time, learning, adapting, smiling.

The optimism is unbreakable.
The future is assumed.

That song, as earnest as it sounds, wasn’t written as propaganda. It was a declaration of faith.

Walt believed automation could end drudgery, not work.
He believed that progress was supposed to buy people back their hours — to relieve them from the mechanical and monotonous so they could spend more of life on what makes it worth living.

Machines, in his mind, were meant to be servants of imagination, not substitutes for it.

He was a capitalist and a complicated one — anti-union, obsessive, ruthless about his standards — but he understood reciprocity. His dream depended on a middle class healthy enough to afford leisure, to take a week off, to buy a ticket to the future he was selling.

He knew you can’t sell wonder to people who can’t afford rest.

That’s what made Carousel of Progress so powerful — it wasn’t just optimism, it was also a promise.

Progress was circular: technology lightened the load, people had time to dream, those dreams created new industries, and the circle began again.

And audiences back then understood that promise. They didn’t fear the machinery; they believed it was there to serve them.

You can see that in Pirates of the Caribbean, the other attraction Walt was building near the end of his life. Those animatronic pirates — rowdy, singing, endlessly repeating their scenes — weren’t a cost-cutting trick.

Walt didn’t automate because he didn’t want to pay performers; he automated because he knew no human could keep that rhythm forever. The actors would tire, their voices would crack, the energy would fade.

The pirates were never meant to replace people — they were built to hold the repetition so that people could stay inventive.

Automation, in Walt’s hands, was an endurance strategy for creativity.

Nobody complained that the pirates had been replaced.
Everyone understood the difference between drudgery and artistry, between the endless motion of work and the human spark of making something new.

That’s the part we’ve lost.

Somewhere between the optimism of Carousel and the automation of today, progress stopped circling back.

We kept the efficiency but forgot the mercy.
We built machines to save time, and then we spent that saved time measuring profit instead of restoring people.

Now AI arrives with the same pitch Walt once sang: shorter workdays, more creativity, a great big beautiful tomorrow.
But shorter for whom? For the worker, or for the owner?
And if it’s for the worker, are they still paid for five?

That’s the real test of progress — whether the prosperity returns to the people who made it possible.

Walt would have understood that immediately.
He wasn’t afraid of automation; he was afraid of forgetting what it was for.

The problem isn’t the technology.

Left unguided, automation follows gravity — toward consolidation, cost reduction, and the quiet erosion of its own audience. Guided with intent, it can still do what Walt meant it to do: carry the weight so people can carry the meaning.

What’s different now is who holds the tools.

For the first time, automation isn’t locked behind corporations and capital. The machinery is no longer hidden inside studios, factories, or boardrooms.

Ordinary people can touch it.

The abstractions are thinning. The systems are learnable. You don’t need permission to design work that removes drudgery instead of removing people.

But that also means the responsibility has shifted.

If automation erases dignity now, it isn’t because it had to.
It’s because someone chose not to learn how to wield it with care.

Learning these systems isn’t technical ambition anymore.
It’s literacy.

If you don’t understand the tools shaping work, someone else will decide what they’re for — and they will optimize for outcomes, not humans.

That’s not a future imposed on us.
It’s one we opt into by disengaging.

When I hear that song now — “There’s a great big beautiful tomorrow, shining at the end of every day” — it doesn’t sound naïve to me. It sounds like a design constraint.

A reminder that progress doesn’t justify itself.
It has to be aimed.

Automation will keep advancing whether we participate or not. The only real choice left is whether we learn enough to guide it — or leave the compass to people who only measure cost.

Walt didn’t build machines because he trusted technology.

He built them because he trusted people — if they were given back their time.

That’s still the bet.

And it’s one worth learning how to make again.

Not sure how AI actually fits into your business?

We offer a focused AI business readiness consultation to evaluate where AI can safely create leverage — and where it’s more likely to introduce risk or noise.

Normally a $500 engagement.
Currently offered at $250.

→ Get the AI Business Readiness Consultation

Everyone keeps asking the same AI question:

“Which one should I use?”

ChatGPT. Claude. Grok. Whatever just showed up in the news this week.

The assumption underneath that question is that there’s a right answer—and that if you pick the wrong one, you’re somehow falling behind.

For most businesses, that framing is already wrong.

AI Is Not a Competitive Sport for Most Businesses

If you run a real company—services, professional work, trades, real estate, accounting, legal, construction, client services—AI is not something you “win.”

It’s not a leaderboard.
It’s not a badge of sophistication.
And it’s not something you should be chasing release by release.

What you’re actually deciding is simpler—and more consequential.

You’re deciding what you want your business to depend on.

The Part No One Wants to Say Out Loud

Most business owners don’t actually want AI strategy.

They want relief.

They want:

• fewer things slipping through the cracks

• fewer follow-ups that only happen because someone remembers

• less mental overhead tracking work

• fewer mistakes caught only by double-checking

AI gets pulled into the conversation because it sounds like leverage.

Sometimes it is.

But here’s the uncomfortable truth:

If your operation already relies on people “keeping an eye on things,” AI does not fix that.

It accelerates whatever system you already have.

Loose systems get looser—faster.
Unclear systems scale confusion.
Manual oversight becomes automated oversight.

That’s why so many people try AI, feel excited for two weeks, and then quietly stop using it.

The problem was the business, not the model.

Why the AI Arms Race Isn’t Your Job

The major AI models are locked in an arms race.

They leapfrog each other constantly.
What looks “better” today will be matched or beaten quickly.

The differences people argue about online—benchmarks, edge cases, micro-advantages—rarely show up in normal business use.

Unless AI is literally what you sell, obsessing over which model is smartest is like a contractor obsessing over which brand of calculator has the best buttons.

It’s not the work.

What actually matters is the company behind the model:

• Will they still be around?

• Will pricing suddenly change?

• Will access tighten?

• Will behavior shift underneath workflows you’ve already built?

That’s the real risk.
Not output quality.

Personality Problems Aren’t Intelligence Problems

A lot of people say they’re “done with ChatGPT” because it’s annoying.

Too agreeable.
Too chatty.
Too much “you’re right to call this out.”

That’s fair.

But it’s not a sign the model is dumb.
It’s a sign you’re seeing defaults.

Most people never adjust instructions or constrain tone. They judge the entire system by the voice it speaks in.

That voice will change.

What doesn’t change is whether your business has a place for the output to land.

If this is useful, subscribe.

If it’s helpful to you, subscribing tells us to keep going.

The Question You Should Be Asking Instead

Instead of asking:

“Which AI should I use?”

Ask this:

What part of my business only works because someone is watching it?

If the answer is “a lot,” then AI isn’t your bottleneck.

Structure is.

AI works when there’s something clear to plug into:

• a defined workflow

• a known handoff

• explicit ownership

Without that, it becomes just another layer on top of the same manual oversight you already had.

That’s where most disappointment comes from.

How We Think About This Ourselves

In our own work, we’ve invested heavily in one ecosystem—not because it’s perfect, but because it was stable enough to build around at the time.

We monitor whether that continues to be true.
If it stops being true, we’ll change course.

That’s not brand loyalty.
That’s operational judgment.

The goal isn’t to pick right forever.
It’s to avoid anchoring your business to something fragile without realizing it.

If You’re Just Poking Around

If you’re early or experimenting:

• pick one

• use it lightly

• learn where it helps and where it breaks

Don’t confuse experimentation with progress.

But if you’re trying to use AI to actually run parts of your business—intake, follow-ups, approvals, handoffs—the model matters far less than whether the system underneath it can hold weight.

That’s the part most people skip.

And it’s why so many “AI initiatives” quietly die after the novelty wears off.

Not sure how AI actually fits into your business?

We offer a focused AI business readiness consultation to evaluate where AI can safely create leverage — and where it’s more likely to introduce risk or noise.

Normally a $500 engagement.
Currently offered at $250.

→ Get the AI Business Readiness Consultation

Each week in , we show how operators design, build, and run systems — from architecture and workflows to the decisions behind them.

Subscribe now

Using AI to Compete in a Crowded Marketplace

If you’re reading this, you likely already use AI in your work. You’re probably above average compared to the general public. But that pool is getting crowded, and the advantage now comes from understanding how to use the tools at a higher level than the people around you.

We’re going to walk through the actual method we use to build productive, stable AI projects. Not hacks, not prompt packs, not hype. Just the mechanics that let you get more out of the tools than most people ever will.

This piece teaches the most important mental shift you can make — something you can apply immediately, today, before you learn anything else.

The First Win: Stop Approaching AI Like Deterministic Software

Most people assume AI behaves like traditional computing:
input → output → repeat.

That assumption is wrong.

Language models are probabilistic. They adapt to context, tone, structure, and intent. They infer far more than they’re explicitly told. And the place where they make the biggest inference — the place where they decide what the entire session is — is the first message you send.

If you start casually, the model becomes casual.
If you start vaguely, the model becomes vague.
If you start with unstructured language, the model will invent structure — usually the generic kind you’ve seen everywhere and dislike immediately.

It’s the result of the environment you unknowingly created.

AI Is Not Just for Coding

A lot of people limit themselves because they assume AI is mainly for code generation or quick rewriting. That belief shrinks what they attempt. But AI is useful in any domain where structured thinking, explanation, summarization, comparison, reframing, or diagnostic reasoning are helpful.

• Operations

• Marketing

• Brand development

• Market research

• Documentation

• Strategy

• Product articulation

• User research

• Internal communication

• Conflict de-escalation

• Workflow design

The list isn’t the point.
The point is that the model’s usefulness extends to anywhere you think, not just where you type syntax.

The reach of the model is large.
How much of it you access depends entirely on how you begin.

Once you understand how that framing works, AI moves from being a novelty to being a genuine part of your workflow.

Why the First Message Is the Environment

The model does not interpret your first message as a question. It interprets it as orientation. It uses the opening to determine:

• the type of work being done

• the tone and role it should assume

• the expected depth

• the level of formality

• the acceptable deviation from user phrasing

• the structure (or lack of structure) it should use going forward

This is why the early versions of this essay sounded generic.
We opened a fresh project with no system instructions, no voice print, and no constraints. The model filled the gap with broad, average English. Once we corrected the environment, the writing shifted.

How ChatGPT 5 / 5.1 Interpret Input

With Chat 5 and 5.1 specifically, these models weight the start and the end of your message most heavily.

• The opening defines the kind of work being done.

• The closing defines the constraints that override everything else.This is why we end messages with structured context blocks wrapped in pseudo-XML:

<context>
  [instructions, constraints, references]
</context>

It cleanly separates the operational content from the conversational content. The model respects what’s inside the block because structurally it cannot mistake it for prose.

This isn’t a trick. It fits how these models actually process input.

Why Prompt Packs Don’t Solve This

Static prompts degrade with each model update.
They work until the next release and then become misaligned.

More importantly, they don’t teach you how to shape the environment.
Without that skill, you’ll always get the model’s defaults — the very patterns you’re trying to avoid.

Everything in these tools depends on one thing you control:

the room you put the model in.

The Mental Model Shift: Your First Message Is the Project

Don’t start a session by asking AI for output.
Start by telling it what kind of work you are doing and what environment you expect it to operate in.

That single adjustment will improve your results immediately, even before you know anything about system instructions, voice prints, or structured reasoning.

The opening message is the part most people overlook.
It’s also the part that determines the entire trajectory of the work.

A Practical Example: Removing “AI-Generic” Through Environment Fixes

Here’s a real example pulled from the earlier drafts of this essay.

What the model originally produced:

“This is not a failure of the model. It’s the result of the environment you unknowingly created.”

Structurally, this is accurate.
But the phrasing uses an extremely common AI pattern:
X is not the problem — Y is.

You’ve seen this everywhere. It’s overused, generic, and easy for a human reader to detect as AI-written. Even though the underlying idea is correct, the construction makes the writing feel automated.

The corrected version:

“It’s the result of the environment you unknowingly created.”

One line removed.
Nothing lost.
Everything reads cleaner.

This is how you challenge the model’s defaults:
you locate the pattern you don’t want, you tell the model what specifically violated your standard, and you ask it to rewrite within the constraints of your voice.

We did this again with another pair of lines:

Original:

“The limit is not the model’s capability. The limit is the initial framing you give it.”

Again: correct idea, generic structure.

Corrected:

“The model is capable of far more than most people attempt. What you get depends on the framing you start with.”

Same meaning.
No X/Y contrarian scaffold.
Human, not synthetic.

The difference isn’t rhetorical.
The early drafts lacked constraints and voice definition, so the model reached for statistically common constructions. Once we defined the boundaries, the model stopped reflexively using those patterns.

This is your immediate win — right now, today.
Start training yourself to identify these generic constructions, and you will be able to shape the model’s output far more precisely.

“AI Slop” and Why This Work Matters

You’ve probably seen the term “AI slop.” The criticism isn’t wrong. You can identify the patterns once you’ve worked with these models long enough: the flattening, the repetition, the softened language, the indistinct phrasing. Most people are producing it because they never pressure the tool and they never pressure themselves. They copy a block of output, paste it into their workflow, and call it done.

That isn’t how this works.

Using AI does not remove the need for effort. You still have to read what it produces. You still have to look for drift. You still have to identify the artifacts that make the writing feel automated. The work is in those small corrections. That’s what gives the output a human edge. You can’t skip that step and expect to stand out.

If you want to differentiate yourself in a market that is now full of people using AI, this is how you do it:
you build the environment, you challenge the model, you check the output, and you refine. That is the gap between people who plateau with the tool and people who move ahead with it.

Whether you’re trying to grow a business, compete in a tighter talent market, expand into new roles, or increase your capacity, the advantage comes from how well you use the tool — not from the fact that you use it.

Start here.
Build the right mental model.
Put in the work that everyone else avoids.
The people who don’t will fall behind.

Share

Leave a comment

Each week in , we show how operators design, build, and run systems — from architecture and workflows to the decisions behind them.”

Subscribe now

“Sameness—done right—moves faster than balance ever could.”

I used to think founders needed opposites.
The visionary and the executor. The dreamer and the realist.
Every book, podcast, and accelerator pitch said the same thing: find your complement.

Then I met my now business partner, Craig—and realized sameness, done right, moves faster than balance ever could.

We think at the same speed. We obsess over the same details. We catch each other’s typos mid-sentence and finish design notes before the other person hits send.

On paper, that sounds dangerous: two people with identical blind spots steering the same ship.

But sameness only kills you if you stop evaluating.
We never do.

The Myth of “Balance”

Balance photographs well on founder decks.
But most balance is translation.

“Balance” sounds responsible. Investors love it.
But most so-called balance is translation. One person speaks in instinct, the other in spreadsheets—and both burn half their energy converting ideas back and forth.

That gap costs time—momentum’s most fragile resource.

Sameness compresses that gap.
When both people process decisions through the same rhythm, you don’t spend energy convincing each other why; you spend it improving the how.

You start the iteration loop faster.
That’s the real advantage. Not yin-yang mythology—velocity.

Our Operating Loop

We run a simple loop:

• Decide. Move while the signal’s fresh.

• Plan. Name the risk, name the resource.

• Execute. Build, test, publish, ship.

• Evaluate. Ask, What did we learn?

• Apply. Adjust the system, not the person.

Then repeat.

“Every launch, every misread, every near-miss becomes data—not drama.”

That feedback loop is our partnership’s real infrastructure.
It’s what keeps sameness from collapsing into echo.

Shared Pace ≠ Groupthink

People assume two similar founders will reinforce bad habits.
Sometimes that’s true. But similarity only creates danger when ego outruns structure.

Our guardrail is ruthless reflection.
We say the quiet parts out loud:
Did we rush that? Did we over-design it? Did we miss the human part?

The conversation that would be awkward in most teams is routine for us.
We built the habit early, when it was just survival—two people figuring out how to ship before the budget ran out.
We couldn’t afford politeness.

Now it’s the operating culture.
Honesty before harmony.
That’s how sameness stays sharp.

Why “Balance” Can Slow You Down

“You can test ideas without being opposites. You just need equal stakes and equal speed.”

Opposite-type partnerships often mistake tension for productivity.
They believe friction keeps ideas tested.
Sometimes it does.
But often it just keeps them tired.

You can test ideas without being opposites.
You just need equal stakes and equal speed.

Craig and I both treat momentum as sacred.
When one of us hesitates, the other fills the space—not to dominate, but to protect the flow.
It’s like editing in real time: cut the pause, keep the rhythm, stay in motion.

That rhythm compounds.
Two aligned founders aren’t twice as fast; they’re exponential.
Each iteration multiplies instead of adds.
The work starts to accelerate on its own.

The Cost of Speed

Of course, there’s a downside: sameness can blur boundaries.
When you both care about everything, you risk caring about nothing in particular.

The fix isn’t more structure—it’s sharper structure.

We divide by responsibility, not hierarchy.
I lead narrative and architecture; Craig leads operations and validation.
But every decision crosses both lanes once before it locks.

“If we can hand it to each other at any stage and it still makes sense, the system’s working.”

That single crossover keeps quality consistent without dragging decisions into committee.

Details as Culture

We don’t celebrate speed for its own sake.
We celebrate precision at speed.

Every deck, doc, and deliverable has to look professional because it signals care.
The details are the culture. They’re how the team knows what “done” means.

People sometimes read that as perfectionism. It isn’t.
It’s stewardship.
The moment you treat the details as optional, you start teaching mediocrity.
The company remembers.

So we protect the details the way we protect each other’s time:
fiercely, quietly, every day.

The Line

Partnership isn’t romance and it isn’t balance.
It’s shared standards held at the same pace.
It’s the relief of not having to translate your ambition.
It’s knowing that when you say good enough isn’t, someone else already agrees.

The hardest part of solo work isn’t isolation—it’s momentum decay.
A good partner keeps the loop alive.

If you want the deeper layers of this work — the wiring, configs, edge-case failures, and the full iterations — you can unlock all of it by becoming a member.

And if you’re at the point where you’d rather have this built inside your business instead of trying to replicate it yourself, start with our $5K Starter System:
10 days. One workflow. No contracts. Full ownership.

Subscribe now

Share

Leave a comment

I was in the office this week for one of my longest-running clients.

We’d been on-site with them for years — through every production cycle, every platform change, every reinvention of how the work got done. Together we built systems that made things faster, clearer, calmer. It was a partnership that worked.

Now they’re closing their doors.
A series of mergers, layoffs, and restructures finally caught up.

It’s not sudden, but it still lands heavy when you’re standing in the middle of it.
There’s a procedure to endings in this business — finishing deliveries, archiving the drives, signing the final paperwork, clearing folder permissions. It’s half ritual, half checklist.

You look around the office — empty edit bays, nameplates half peeled off — and realize it’s all winding down.

Being one of the last people left is a strange kind of responsibility.
You remember where everything lives, what every note meant.
You’ve already done your grieving; you made peace with it months ago.
But when it finally hits paper — the signatures, the empty desks — it still feels like a loss.

The Chalkboard

I was in the kitchen grabbing a water when I noticed the chalkboard — still hanging there, mostly blank.

A few faint outlines from old designs were left behind, and it made me think of one of the younger team members who’d started there not that long ago.

He was an intern who’d worked his way up to PA.
Smart, steady, reliable. The kind of person who noticed details most people miss.

He was the one who used to take that chalkboard every month and turn it into something great — birthdays, holidays, little seasonal jokes.
Fun, simple designs that gave the office some personality.

He didn’t do it because anyone asked him to.
He just had the natural impulse to make things a little better — to bring some order and life into the room.
That’s what always stood out to me. You can’t teach that.

It’s the kind of instinct that, if you nurture it, can become the thing you build a career around.

He happened to be there that day, picking up a few last boxes.
We caught up for a bit.

He told me he was moving back home, taking shifts at the diner for a while.
He smiled when he said it — that polite, practiced kind of smile people use when they’re trying to sound okay.

He’ll be fine. I don’t doubt that.
But I could see the hesitation — that mix of fear and logic you only get when you’re young and standing on the edge of something uncertain.

I told him again what I’d said before: that he had a gift, and that people would pay for that kind of work.
Chalkboards, weddings, signage, murals — simple ideas, small ways to start.

He said he’d think about it.

And I get it. I really do.
Stability is a powerful thing.
It’s hard to walk away from the safety of what you know, even when what you know is ending.

Sometimes the best advice you can give someone is the simplest one: just start.

It’s easy to say and hard to live by.

Planning feels responsible. Reflection feels smart.
But if it never turns into movement, it’s just a loop.

The Rhythm That Works

I’ve seen it play out in productions, in business, in myself.
You can’t plan your way to motion.
You have to move your way to better planning.

That’s how I run my business.
It’s how I like to lead my teams.

Plan. Move. Adjust.
Over and over.

I learned that rhythm years ago with a counterpart at the client — a VP who ran the production schedule with me.
Every time we sat down to map it out, we’d only plan two weeks ahead.
Anything beyond that was guesswork.

We’d lock what we could see, then start.
The moment we moved, the plan changed — and that was fine.
That was the point.

You learn while you’re doing.
The plan gets sharper because you’re already in motion.

That’s the same rhythm I keep now with my business partner.
It’s how we run everything — short windows, constant recalibration.

We’ll look two weeks out, maybe three if we have to, tighten what’s in front of us, and keep moving.
That’s what keeps the company alive.

We don’t try to lock the next six months.
We make the best next move, measure it, and make another.

Momentum and Trust

Reflection matters. Thinking matters.
But the work only breathes if you’re doing both at the same time.

That’s what I hope that young guy eventually figures out — that momentum doesn’t mean rushing ahead blindly.
It means keeping the loop alive: plan, act, learn, adjust.

Don’t freeze.
Don’t wait for certainty.

He’s got time. He’ll find it.
But I hope he starts soon.

Because in the end, that’s what keeps any of us working —
not the perfect plan, not the guarantee,
just the willingness to move.

That’s what’s carried me through every closing, every reset, every start.

Plan. Move. Adjust.
That’s the work.

It’s become its own genre:
“100 ChatGPT prompts to change your life.”
Swipe, scroll, bookmark.

Every one of them looks the same:

“Write a business plan for [insert your idea].”
“Generate 10 product names for [insert your product].”
“Draft a pitch email for [insert your audience].”

The trick is hidden in plain sight.
The power isn’t in the prompt.
It’s in the brackets.

The Prompt Hustle

I’ve lost count of how many LinkedIn posts or Twitter threads I’ve read that promise “the ultimate prompt pack.” They rack up likes, comments, and reshares. People really want to believe the magic is in the words you feed the machine.

But look closer. The structure is always the same. The actual prompt is filler. The system is selling you the illusion of leverage.

The real load-bearing part — the thing that makes the AI useful or useless — is what you put in the brackets. Your business idea. Your product. Your story. That’s the part they can’t sell you.

It’s the hardest part. It’s also the only part that matters.

What the Brackets Really Mean

• “Write a sales plan for [my product].”
  → But what is the product? Who buys it? Why would they care?

• “Write a funding pitch for [my startup].”
  → But what problem do you solve? What market are you entering? What proof do you have?

• “Write a marketing strategy for [my brand].”
  → But what’s your positioning? Your wedge? Your voice?

The machine fills space. You still have to fill the brackets with something real.

That’s not a trick. That’s the work.

Why the Brackets Hurt

Most people stumble here.
Because what goes in the brackets isn’t surface-level.
It’s identity.

• Naming the real thing you want to build.

• Accepting the risk that it might fail.

• Distilling the messy human intent into a single sentence that can survive automation.

AI can’t carry that for you. It will carry the repetition, the scaffolding, the formatting. But it will not invent the thing that matters in your brackets.

That’s your job.

The False Shortcut

The prompt packs are popular because they hide that truth. They let you feel like you’re making progress without facing the hard work of defining what belongs inside the brackets.

It’s the same instinct that fills hard drives with half-finished pitch decks, or notebooks with abandoned outlines. Collecting prompts feels like motion. But motion isn’t traction.

At some point, you have to stop scrolling and name your damn product.

What Doing the Work Looks Like

Doing the work isn’t glamorous. It means:

• Interrogating the brackets. Not “a book,” but your book. Not “a candy bar,” but your candy bar, your market, your brand story.

• Breaking the silence. Writing the messy draft, the first outline, the ugly version you’d never show anyone.

• Iterating in public. Testing your idea where it can fail, and learning if anyone cares enough to pay.

AI can draft 10,000 versions. None of them matter until you’ve put something true into the brackets.

A Personal Note

I use AI every day. It carries the load. It structures my thoughts, cleans up my phrasing, and helps me see the edges of my own argument. I’m not anti-prompt.

But I don’t confuse the filler words with the work.
The work is sitting with the bracket until it’s not empty anymore.

That’s why most of the “ultimate prompt packs” feel hollow. They skip past the hard part. They pretend the brackets will magically fill themselves.

They won’t.

The Human Part

Prompts don’t build businesses.
People do.

AI doesn’t save you from naming what you want to make.
It just waits for you to fill the brackets.

And if you can do that — if you can sit with the silence long enough to name your book, your show, your product, your company — then the machine becomes powerful.

Until then, it’s just filler.