The Most Important Job of the Next Decade? Editor.

McKinsey is cutting 10% of its workforce. The firm that tells everyone else how to restructure is restructuring itself.

AI is automating the very work consultants built their model on. The gathering of data, synthesising research, building slide decks and generating first drafts.

That last one matters most.

The first draft used to be expensive as it took time, training and proper graft to get anything onto the page. Now you can generate ten versions of almost anything before your coffee gets cold.

This doesn’t mean creation got easier. It means the bottleneck moved and the hard work now sits at both ends.

Upstream: deciding what’s actually worth making, who it’s for and why it matters.

Downstream: knowing whether what comes back is any good, fixing what counts and standing behind the result.

The draft in the middle? That’s the easy part now.

AI writes. We edit. That’s the new division of labour.

Mind you, “editing” doesn’t mean what most people think it means.

⸻

Above the Line, Below the Line

In book publishing there’s a distinction between what editors do below the line and above the line.

Below the line is what most people imagine, grammar, clarity, consistency, polish, application of red pen, tutting etc.

Above the line is everything else, Should this exist at all? – What is it really trying to say? – What’s missing? – Who is this actually for? – When is it done?

Peter Ginna, editor of What Editors Do, describes the role as being a connector, a conduit between writer and reader, a translator or someone who improves communication in both directions.

That’s not someone fixing commas, it’s someone standing between creation and audience asking one hard question:

Does this work?

Jonathan Karp, now CEO of Simon & Schuster, puts it more bluntly. Editors earn their keep at the acquisitions stage. Choosing what to bet on. “No amount of brilliant editing can turn an unsaleable book into a winner.”

The skill isn’t polish. It’s judgement about what deserves to be polished in the first place.

⸻

What McKinsey Is Really Cutting

When McKinsey talks about its AI strategy, it’s explicit about what stays and what goes.

They’ll keep hiring people who face clients and they’ll shrink the layers that gather, synthesise and present information.

Production work is being automated where judgement, relationships and accountability are being protected.

In other words, the people who decide what to make and whether it worked.

Editors.

This isn’t about one firm or one industry. As one analysis of the cuts put it, “the premium for future talent will no longer rest on analytical horsepower alone.”

The old moat, being good at processing information has drained away.

What’s valuable now is knowing what the information means, whether it matters and what to do about it.

⸻

“Editor” Is Not the Grammar Police

When people hear “editor,” they think red pen.

That’s not the job.

The real job is taste, judgement, and accountability. The ability to say this works or this is nonsense and live with the consequences.

These skills were always valuable. They were just harder to see when we were busy typing.

Publishing figured this out years ago. Editors were never content producers. Manuscripts arrived in huge volumes, most of them unusable. The job was selection, shaping and saying no far more often than yes.

That’s now everyone’s job – The lawyer reviewing AI-drafted contracts – The strategist sifting AI-generated scenarios – The marketer choosing between AI-produced campaigns – The leader deciding which insight to back and which to bin.

The cost of production has collapsed and the value of selection has gone through the roof.

⸻

The Skill That Was Hiding in Plain Sight

For decades, consulting and knowledge work ran on a comforting assumption that the hard part was doing the work. Analysis. Research. Synthesis. Presentation.

Clients paid for output.

AI exposes what was always true, the output was never the point as the real value was knowing what question to ask, recognising the right answer when you saw it and having the nerve to act on it.

Clearly those aren’t analytical skills, they’re editorial ones.

⸻

What Editing Looks Like at Work Now

This is the part most people miss.

Modern editing isn’t about fixing text. It’s about shaping thinking.

The editing moves that matter now: 

• Framing: “What problem are we actually solving here?” 

• Audience editing:“Who is this really for and what will they care about?” 

• Insight extraction: “Which trend or data point matters and which is just noise?

• Assumption testing: “What would have to be true for this to work?” 

• Selection: “If we could only keep one idea, which survives?” – 

• Stopping: “This is good enough. We’re done.”

These show up as prompts too: For example, “What’s the strongest version of this argument and why might it still be wrong?” – “What would a sceptic say in one sentence?” – “What’s missing that would change the decision?” – “If this failed in six months, what would we say we ignored?” – “Which part is trying too hard?”

AI is very good at generating options. It is terrible at choosing.

That’s on us.

⸻

Start Now

At the end of last year we watched a leadership team use AI to generate five versions of a strategy in about ten minutes. Perfectly coherent, nicely structured, all of them very plausible.

Then they sat looking at each other as nobody could say which one was right, or whether any of them were. The AI had done the writing but it couldn’t tell them what they actually believed.

That’s the gap and it’s not going away, so it’s time to train your editorial instinct.

Read more and notice why things work or fall flat. Practice explaining what you’d cut, not just what you’d add. Get comfortable making calls with incomplete information because that’s all you ever have.

Most importantly, get used to being accountable for decisions AI helped you make but won’t help you defend.

The machines can write. But they can’t decide what’s worth writing, or whether it’s good enough. They can’t take responsibility when it matters.

That work has a name. It's called editing. And it's not going anywhere.

Next time: We built a thing. It's a boardroom full of opinionated execs who'll tell you what's wrong with your idea. Except they don't exist, they won’t judge and they're available at 11pm on a Sunday. We'll show you how it works.".

What We’re Reading

Three pieces this week that all circle the same uncomfortable question: in a world where AI can produce anything, who decides what’s actually worth making? The answers point the same direction, toward judgment, discernment and the stubbornly human skill of knowing when to say no:

The Rise of Taste: Why Human Curation Will Define the AI Era — Debris Studio “Taste is a responsibility. It’s not just about what you like. It’s about what you allow in.” A design studio argues that in a world drowning in AI-generated content, the scarcest skill isn’t creation, it’s the wisdom to know what’s worth creating in the first place.

Velocity Is the New Authority. Here’s Why — Om Malik Authority used to be the organising principle of information. You earned attention by being right. That world is gone. Now the algorithm doesn’t care whether something is true, it cares whether it moves. The result: a culture optimised for first takes, not best takes.

AI Is Everywhere. Editors Should Be, Too — Poynter A catalogue of AI-generated disasters from fake books, to fabricated sources, to hallucinated facts, all with one thing in common: no editor in sight.