Picture a busy bar on a Friday night. A machine could pour the drinks faster, cheaper, and to the milliliter. But the bar is packed anyway, and people are leaning in to talk to the bartender. They are not paying for the gin. They are paying for him — the read on the room, the half-remembered story about last week, the face that recognizes them.1
Koe's claim is blunt: the jobs that survive the AI wave are the ones where the human is the product.2 Not what they make. Who they are. The moment the value of the work can be peeled off the person who did it, a machine takes the work. When the value stays welded to the specific person, the person keeps the job — and gets to charge more for it.
Strip away the futurism and this is a sorting rule for work. Ask of any role: can you swap out the human without the buyer noticing or caring? If yes, that role is on the automation list. If no — if the buyer is buying this person specifically — the role holds.2
Koe names four buckets where the human stays welded to the value.3 High-liability roles, because somebody has to be accountable, and you can't sue a model. Statutory and legal positions, where the law requires a named human in the chair. The experience economy — bartenders, boutique shops, art galleries, live performances — which he predicts will boom. And relationship-and-trust roles: sales, diplomacy, negotiation, where the whole transaction runs on a human believing another human.3
The through-line: in each case the deliverable is inseparable from the identity, the body, the reputation, or the legal personhood of the human doing it. [PARAPHRASED] [POPULAR SOURCE]
Three different engines drive the same result, and it helps to see them as separate.
The first engine is accountability. A high-liability role exists because someone has to absorb the blame when it goes wrong. Blame needs a body and a bank account. A model has neither, so the law and the market keep a human in the loop — not because the human is faster, but because the human can be held.3
The second engine is presence. The experience economy sells the fact that a human was there, in the room, doing the thing while you watched. A live performance is valuable partly because it can't be redone — the bartender's improvisation, the gallery owner's eye, the chef's hand. Koe expects this to swell into "a huge boom" precisely as the friction-work drains away to machines.3
The third engine is trust transfer. Sales, diplomacy, negotiation — these run on one human deciding to believe another. You can't outsource the believed-in party. The trust is the product.3
Behind all of it sits Koe's bigger frame, which he borrows from a novel: "There are almost 7 billion people on this planet. Someday I hope there will be almost 7 billion companies."4 [UNVERIFIED] Every person, his read goes, becomes their own enterprise — the human not as employee but as the firm itself. [PARAPHRASED] [POPULAR SOURCE]
This page hands the vault a filter. Drop any skill, role, or product through it and ask: is the human the product, or just the producer? That question routes straight into You Are the Last Defensible Moat, which gives the precise test — swap creator and creation; if value survives the swap, it's automatable.
It also sets up Speed to Escape vs Premium to Savor, its sibling: the experience economy boom Koe predicts is the savor side of that split. And it feeds Reputation as the Most Important Asset — when you are the product, your reputation is the spec sheet.
Koe runs the swap test on two photographs.5 A generic stock photo of someone at a laptop — swappable. Any machine can spit out a thousand. Nobody asks who shot it. Now a photograph by a famous photographer. Not swappable. The value lives in "her eye, her choices, her reputation," in everything about her that shapes how you read the image.5
Same object — a photo — sitting on either side of the line. The pixels aren't the product in the second case. The photographer is. Take her name off it and hand it to you cold, and it doesn't carry the same weight. That's the whole thesis in one frame: the human is welded to the value, and the weld is the moat. [PARAPHRASED] [POPULAR SOURCE]
It's Tuesday afternoon. You've got a notebook open and your own work in front of you — the last three things you shipped. You pick one up and you ask it out loud: if I took my name off this and a stranger handed it to a client, would anyone care?
The first one is a templated report. Cold-handed, it loses nothing. You feel the floor drop a little. That's the swappable pile.
You pick up the second — a proposal you wrote after two phone calls where the client told you things he hadn't told anyone. You try the swap. It falls apart. The value was never the document; it was that you were the one he trusted enough to say it to. You set that one in the keep pile.
By 4:30 you've sorted everything into two stacks. The swappable stack is your automation roadmap — hand it to the machines on purpose. The welded stack is where your hours should go from here. You're not predicting the future of work. You're auditing which half of your own week already lives in it.
You're at risk of being on the swappable side when:
The "7 billion companies" line is attributed by Koe only to "a novel that I read a long time ago" — he names no title or author.4 🚩 SINGLE SOURCE, and the attribution is unverifiable as stated. [UNVERIFIED]
The experience-economy "boom" is a prediction, not a measured trend.3 🚩 MOTIVATED REASONING — Koe sells creator-economy education, so a future where the human-as-product wins is a future where his product is the on-ramp. Document, don't endorse.
Open tension: Koe lists high-liability and statutory roles alongside bartenders and artists, but these survive for opposite reasons — one by legal compulsion, the other by buyer desire. Lumping them under "the human is the product" is tidy but flattens a real difference. The lawyer-in-the-loop may be a temporary regulatory artifact; the savored experience may be durable.
Koe and the Perell positioning corpus are looking at the same shift from two angles. Perell's Everybody Is a Media Company says the sovereign individual broadcasts and accrues an audience; Koe's "7 billion companies" says the individual becomes an enterprise. Where Perell emphasizes the media layer — you publish, therefore you have leverage — Koe pushes further into the experience and trust roles where there's nothing to publish at all. The bartender isn't a media company. He's the product without ever posting. That's the seam between them: Perell's frame covers the visible creators; Koe's stretches to cover the people whose product never goes online.
He converges hard with Jack Moses on Reputation as the Most Important Asset. Moses argues reputation is the one asset that compounds and can't be bought; Koe is describing the mechanism by which it becomes the only asset that matters — once the output commoditizes, the buyer is choosing the reputation, full stop. They don't clash; Koe supplies the macro reason Moses's claim sharpens over time.
Plain version: a famous photographer and a trusted salesperson are doing the same trick — they've made themselves un-swappable — and seeing that they're the same trick tells you what to actually build.
First handshake: Perell — Everybody Is a Media Company (behavioral-mechanics). Perell's frame is about distribution leverage: own an audience and you own a channel nobody can revoke. Koe's "human is the product" frame is about substitution resistance: be the thing buyers can't swap out. These look identical until you press them. A media company can be swapped — audiences migrate, channels die, the next creator does the same format better. But a human welded to a high-trust or high-liability role is harder to migrate away from, because the trust took years and the liability needs a body. The insight that falls out of holding both: distribution leverage and substitution resistance are different moats, and the creator who only builds distribution has a moat that drains. The bartender has no distribution and a deep substitution moat; the viral poster has huge distribution and a shallow one. The durable position is the overlap — be the followed person and the welded-in one. Neither Perell's media frame nor Koe's product frame, taken alone, tells you to chase both at once; lining them up does.
Second handshake: Reputation as the Most Important Asset (business). Moses treats reputation as an asset class — something you bank and compound. Koe's page explains the demand-side force that inflates that asset's price: as production commoditizes, buyers have nothing left to discriminate on except who. Put the two together and reputation stops being one asset among several and becomes the residual claimant on all value — the only thing left to charge for after the machines have taken the work. That reframes a lot of "personal brand" advice. It's not vanity; it's the literal product. When the human is the product, reputation is the price tag, the warranty, and the spec sheet at once — a synthesis neither the asset view nor the survival view reaches on its own.
The Sharpest Implication. If the human is the product, then "improving your output" is, past a point, the wrong investment — you're polishing the swappable part. The leverage is in deepening the weld: the trust, the trajectory, the accountability, the presence that can't be peeled off you. Most professional advice optimizes the exact layer that's about to commoditize.
Generative Questions.