Most "skills you need in the AI age" articles read like a grocery list: learn marketing, learn sales, learn writing, learn to prompt. Dan Koe says that's the wrong shape entirely. The skills you need aren't items in a basket — they're rungs on a ladder, each one resting on the one below it.1 You can't really climb to the second rung until you're standing on the first, and the order is not negotiable.
Here's his ladder, bottom to top: agency, then taste, then perspective, then persuasion, then technical know-how.2 The counterintuitive part is which way the ladder points. Most people start at the top — they rush to learn the tool, the software, the AI — and wonder why nothing compounds. Koe says the tool is the last rung, the least important, the most easily replaced. The rung that holds the whole thing up is the one nobody can teach you in a tutorial: agency.
Strip the framing and here's the claim. These five aren't career-specific competencies, because in Koe's future you aren't entering a specific career.3 They're human skills — closer to capacities you uncover than subjects you study. The exercise, he says, is less "acquire these" and more "uncover what your nature is and lean further into it."4
The word doing the heavy lifting is hierarchy. The five umbrella under one another: agency makes taste possible, taste shapes perspective, perspective fuels persuasion, persuasion directs the technical tools.5 Read the wrong direction and it falls apart — learning a software package teaches you nothing about what's worth making, and learning to make people care means nothing if you have no perspective worth caring about.
The order also encodes a defense. The higher you sit on the stack, the harder you are to automate, because the lower rungs (technical know-how especially) are exactly what machines are eating.6 So the hierarchy is simultaneously a learning sequence and a ranking of how AI-proof each layer is, with the most human capacity — agency — sitting safely at the bottom holding everything up.
Think about why a ladder works at all. The top rung is the one you stand on to do the thing — reach the gutter, pick the fruit. But the top rung is useless without the bottom rung bearing the load. Koe's stack is built the same way. Technical know-how is where the visible work happens — you actually open the software and make the thing. But it bears no weight on its own. Without agency underneath, you're just executing someone else's instructions with a fancier tool.
Now flip it to see the automation logic. Machines climb the ladder from the bottom of the visible work — they took manual labor, then animal labor, then machine labor, now cognitive labor.7 Each wave abstracts the technical layer up and out of human hands. So if you've staked your value on the technical rung — knowing the software, doing the manual production — you're standing on the rung the machine reaches first. The only durable position is higher up, on the rungs that require you to exist: a self that chooses (agency), discerns (taste), sees (perspective), and moves others (persuasion).
This is why Koe insists it's a stack and not a list. A list lets you grab the easy items and skip the hard one. The hierarchy won't let you — skip agency and the four rungs above it have nothing to rest on.
The vault holds a lot of individual skill claims — persuasion, taste, positioning, technical tooling — often floating without an organizing order. This page hands them a backbone. It tells you that Persuasion as the Greatest Skill is real but fourth — it only pays off when there's a perspective worth persuading people toward, which is a correction to any page that treats persuasion as the master move. It positions the whole AI-tooling cluster, like Taste/Judgment/Labor, as the bottom rung, the most replaceable, which reframes "learn the tools" as necessary but lowest-priority.
It also nests the sibling pages cleanly: Agency as the Meta-Skill is the foundation this stack rests on, and Taste as Curation Over Creation is rung two. And it gives Skills Abstract Up a Layer its practical payoff — if skills keep abstracting upward, the stack tells you which direction to climb to stay ahead of the abstraction.
Koe's own example for why persuasion sits where it does is the buried book. Picture an author who works ten years on an insane, genuinely great book — and then nobody sees it. It's worthless, Koe says flatly, not because it's bad but because the persuasion rung was missing.8
Deconstruct what went wrong in stack terms. This author clearly had the lower rungs — agency (they chose the decade-long project), taste, perspective (the book is good). What they lacked was rung four: the ability to make people care.9 The case is instructive precisely because it's not a story about a hack who skipped the work — it's about someone who did everything except the rung that converts a finished thing into a thing that reaches anyone. Koe's prescription is concrete: to build that rung, learn marketing, sales, copywriting, and build on social media — and learn them the only way that works, by doing them, making mistakes, noticing what isn't working, trying new techniques, getting better.10 The buried book is the cautionary inverse of the whole stack: a perfect lower three rungs with the fourth sawn off, and the thing falls.
It's Sunday evening. You've got a notebook open and you're doing something unusual — auditing yourself one rung at a time, bottom up. Rung one, agency: you ask, honestly, am I choosing my own problems this week, or waiting to be told? You write down one place you've been waiting. Rung two, taste: you scroll your own recent work and ask which pieces you actually stand behind versus which you posted to fill the quota. You feel the difference and mark it.
You keep climbing. Perspective — are you holding any genuine complexity, or just repeating a position you absorbed? Persuasion — does any of this reach people, or is it a buried book? Technical — and only now, last, do you ask about tools, and you notice you'd been spending most of your worry here, at the top, on the rung that matters least. You close the notebook with one correction: this week you'll spend your attention low on the ladder, on choosing and discerning, and let the tools be the easy part they actually are.
Koe presents the exact five-rung order as if it's settled, but the ordering is an assertion, not an argument.11 [PARAPHRASED] [POPULAR SOURCE] You could plausibly argue perspective precedes taste rather than following it — how do you discern without a vantage to discern from? Koe doesn't defend the sequence rigorously; he states it. Hold the direction (human capacities below, tools on top) as the strong claim and the precise middle ordering as looser.
🚩 MOTIVATED REASONING — the stack conveniently ranks "learn the specific tool" (which dates fast and is hard to sell as evergreen) below "develop yourself" (which Koe's courses teach and which never expires). Note the alignment between the hierarchy and what's monetizable for him. Open question the source leaves dangling: the stack says technical know-how is least important and most replaceable, yet the entire video is bracketed by a pitch for a specific tool (Eden).12 That's a live tension — if the tool layer matters least, why does the talk open and close on tooling? The honest reading is that tools are low-priority to master but still necessary to use, and Koe blurs that line where it serves the pitch.
Koe and Butcher converge hard on the bottom two rungs. Butcher's Personal Monopoly 3C Framework is essentially a route up Koe's lower stack — building a defensible position out of who you are, which requires agency and taste before anything else. Where they'd part: Butcher tends to frame the climb as a constructive project (build your monopoly), while Koe frames it as uncovering a nature already there. Build versus reveal — a real difference in how the lower rungs get developed.
The convergence with Hormozi/NEPQ comes at rung four. The sales pages in the vault — like Sales as Discovery Not Persuasion — would actually correct Koe's word choice. Koe labels rung four "persuasion"; NEPQ insists persuasion-as-pushing fails and that the real skill is discovery that lets the prospect persuade themselves. Read together, Koe's fourth rung is better understood as the skill of making people care — which NEPQ would say is achieved through questions and discovery, not through persuasion in the pushy sense. The vault's sales corpus sharpens the rung Koe names most crudely.
Plain version: this five-rung ladder of human skills is the same shape as how a craftsperson actually develops, and it lines up almost exactly with how the creativity pages describe building real creative capacity — which means the "AI skill stack" is not new advice at all, it's old developmental wisdom wearing a 2026 hat.
The first handshake is with Butcher — Personal Monopoly 3C Framework in creative-practice. Butcher's framework builds a defensible creative position from clarity, consistency, and a compounding body of work. Lay it against Koe's stack and a structural parallel appears: both are bottom-up developmental sequences where the durable, valuable layer is the self-built foundation and the visible output sits on top. But hold them together and something appears that neither states alone — Butcher's "personal monopoly" is precisely what a fully-climbed Koe stack creates. Agency chooses the territory, taste and perspective make it distinctive, persuasion makes it reach people, and the tools execute it; the output is a monopoly of one. Koe gives Butcher the learning order; Butcher gives Koe the end-state name for what the climb produces. Either page alone is half a map; together they're a full route from "I have no defensible position" to "I am the only one who can occupy this."
The second handshake is with Taste/Judgment/Labor in the AI-collaboration cluster. That framework splits creative work into three layers and asks which layer AI takes (labor) and which stays human (taste, judgment). Koe's stack is a more granular version of the same insight, and overlaying them is clarifying. Taste/Judgment/Labor's "labor" maps to Koe's bottom rung (technical know-how, the part AI eats); its "taste" and "judgment" map up Koe's stack to taste, perspective, and the discernment inside agency. The insight that emerges from holding both: the vertical axis is an automation gradient. As you climb either framework, you move from what machines do to what only humans do — which means a creator's career strategy is literally a climbing instruction: spend your scarcest attention as high as you can, because every rung down is a rung closer to being automated. Two frameworks, one underlying gradient, and the strategy falls out of the overlay rather than from either page by itself.
A third, lighter connection runs to Skills Abstract Up a Layer. If skills keep abstracting upward over history, the stack answers the natural follow-up — up toward what? — by naming the ceiling: agency, the capacity to act without permission, the one rung that can't abstract further because there's nothing above "deciding for yourself."
The Sharpest Implication. If the stack is real, then almost everyone is optimizing the wrong rung. The market for "learn this AI tool" is enormous because the tool layer feels concrete and teachable; the market for "develop agency" is small because it's uncomfortable and unsellable as a quick win. So the crowd rushes the top rung — the replaceable one — and leaves the bottom rung, the durable one, uncrowded. The stack is, in effect, a contrarian allocation signal: invest where the crowd won't, low on the ladder.
Generative Questions.