essay
Part 7: What Is Expertise When AI Can Express Any Expert's Frames?
If you’d asked this question fifteen years ago, the answer would have been simple.
Experts were the people who could produce things others couldn’t.
The strategist who saw what others missed. The doctor who knew the rare diagnosis. The lawyer who understood the case law in their bones. The engineer who could solve the problem.
Production was the differentiator. The artifacts proved the expertise.
That is not what experts are anymore.
AI can produce a strategy. AI can produce a legal analysis. AI can produce a credible diagnosis from a list of symptoms. AI can produce an engineering plan from a problem brief. AI can produce a market thesis from public data.
The artifacts no longer prove anything specific. They prove access to the tool. Not the expertise behind them.
So what is expertise now?
When the artifacts are commodity, what’s left?
There Are Three Types of Operators in This New World
There are operators who are curators. Their work is choosing.
There are operators who are still producing. Their work is being commoditized in real time.
And there are operators who don’t know there’s been a shift.
In five years, only the curators have a defensible practice.
The other two are competing for a floor that keeps falling.
OPERATOR FILE #21
Expert operators know the discrimination is the expertise.
Average operators are still producing.
Commodity operators don’t know the difference.
The answer the world is slowly converging toward looks like this:
Expertise is the discrimination that decides which of an infinite supply of plausible answers is the right one for this specific situation.
It is not the production. It is the curation.
A curator at a museum doesn’t paint. The painters paint. The curator selects.
Out of millions of paintings, billions of artifacts, an enormous infinity of possibility — the curator selects the precise set that should hang on the wall for this audience, in this season, in this room.
The curator’s expertise is invisible if you only watch the production.
It becomes visible the moment you ask: out of everything possible, why this?
That question is what AI cannot answer for you.
The model can produce a thousand strategic frameworks. It cannot tell you which framework is right for this client, given this market, this leadership team, this moment.
The model can produce a hundred technical solutions. It cannot tell you which one will compound — versus which one is a trap that looks the same now and won’t in eighteen months.
Choosing IS the expertise.
Producing is now free.
OPERATOR FILE #22
Expert operators have flipped their time allocation. Most of their hours are choosing.
Average operators are still spending most of their hours producing.
Commodity operators don’t measure the ratio.
This reframe changes what experts spend their time on. If they take it seriously.
The old time allocation:
Most of your hours producing artifacts. A small fraction deciding what to produce.
The new time allocation:
A small fraction of your hours producing — the model handles most of that. Most of your time deciding what should be produced. And reading whether what was produced is actually right for the situation.
The old skill: production excellence.
The new skill: discrimination at the level production can’t reach.
Most knowledge entrepreneurs are still operating on the old allocation. Their week is full of producing things. They are using AI to speed up the production. Their discrimination — the part that’s actually their asset now — gets practiced in the cracks between productions, when there’s time. Which is rarely.
The flip is uncomfortable.
It feels like you’re doing less work.
You are.
The work that’s left is the part that used to fill twenty percent of your time. Now it should fill seventy.
You’ll feel less productive.
You’ll be more valuable.
The Test
Pick an engagement you finished in the last six months.
Look at what was actually load-bearing.
Was it the deliverable? The deck. The report. The analysis. If yes — that work is now a commodity. The price will compress over the next two years.
Was it the call? The diagnosis you made about what the client actually needed. The strategic recommendation that was different from what the brief asked for. The objection you raised in the kickoff meeting that changed the entire scope.
If yes — that’s the expertise. That is what’s getting more valuable. Not less.
Most engagements have both. The ratio between them is the question worth holding.
If your engagements are 80% producing and 20% calling, you’re working on the wrong things for the world we’re now in.
If your engagements are 20% producing and 80% calling — and the 20% production is the tactical work the call required — you’ve made the shift.
OPERATOR FILE #23
Expert operators get better by choosing more.
Average operators get better by producing more.
Commodity operators don’t realize the rep type matters.
Here is the part most experts miss.
The curator’s expertise grows differently than the producer’s.
Producers get better by producing more. More reps. More outputs. More practice.
Curators get better by choosing more. More discriminating moments. More decisions. More situations where they had to read what the situation required.
AI can give you infinite production reps without making you a better producer.
That is the trap the rest of this series has named.
What’s harder to see is that AI can also give you the opportunity for more curating — if you use it deliberately.
Every time the model produces three options and you have to choose between them — that is a discrimination rep.
Every time it produces a framework and you have to decide whether it fits this client — that is a discrimination rep.
Every time it gives you a confident answer and you have to decide whether to trust it — that is a discrimination rep.
The same tool that hollows expertise out of one operator can sharpen it in another.
The variable is whether the operator treats every interaction as production (atrophy) or as choosing (compounding).
What Was Always Underneath
Expertise was never the artifacts.
We thought it was for a while because, until recently, only experts could produce the artifacts. The artifacts were a reliable proxy for the expertise underneath.
That proxy has broken.
The artifacts no longer signal anything specific.
The expertise — the discrimination, the choosing, the curation — is what was always under the artifacts.
And it is the only part that’s getting more valuable.
The work for the next decade is to spend your time on it deliberately.
The market hasn’t figured this out yet.
It will.
Slowly, then quickly. The way these things always do.
The operators who make the shift first will be the ones the market eventually pays its highest premiums to.
Declaration
This is the synthesis. Six pieces. Six installs.
Say them all. One hand on your heart:
“I am a judgment-builder.
I protect what I’ve built.
I do the slow work in new domains.
I know where my real judgment ends.
I protect the space where it forms.
I price for discrimination, not speed.”
Now put both hands on your heart and say — out loud:
“The output is borrowed. The choosing is mine.”
Say it again. Mean it.
“The output is borrowed. The choosing is mine.”
That is the entire architecture. You are standing in it now.
Action
This is the last action of the public series. Take it slowly.
Pick the engagement on your calendar that matters most this quarter.
Open a blank document.
Write down — in order — every choice that engagement requires you to make. Not the deliverables. The choices. The diagnostic calls. The strategic decisions. The objections you’ll need to raise. The places where you’ll have to discriminate between options that look equally plausible.
That list is the work.
The deliverables are the residue.
When you can see the difference clearly, the rest of your career bends in a different direction.
The output is borrowed.
The choosing is yours.
Make sure it stays that way.