The unit of capability is changing — Editorial Intelligence

The interesting question is not how many content tasks AI can automate. It is what happens to the design of a content organisation when every talented person inside it becomes much more capable.

I think we may be underestimating what AI does to the shape of a content team.

Most of the discussion still starts with tasks.

Can AI write the article?

Can it summarise the research?

Can it generate social posts?

Can it make the image?

Those are reasonable questions, and they have increasingly obvious answers. But the more interesting change happens when those capabilities start to compound.

An experienced content marketer can increasingly move across work that used to require several different handoffs.

Research becomes easier to interrogate. Existing knowledge becomes easier to retrieve. Arguments can be tested quickly. Drafts become cheap. Editing becomes iterative. Different formats become easier to explore. Basic creative production becomes accessible. Analysis can feed directly back into the next decision.

None of those things individually replaces a content team.

Put them together, though, and something changes.

The capability of the individual expands.

A strong editorial operator with AI starts to look less like a person completing one part of a production line, and more like someone capable of orchestrating a substantial part of the system.

The uncomfortable version

That has an implication worth stating plainly.

The big effect of AI on content may not be that machines become good writers.

It may be that good content people become extraordinarily capable.

Which is a different claim, and a more disruptive one, because it doesn’t depend on the technology being better than a professional at anything. It only depends on the cost of executing each individual task falling far enough that one person can credibly hold more of the process at once.

”Productive” is the wrong word

I don’t mean productive as in producing fifty pieces instead of ten.

That would be the least interesting outcome available, and probably a harmful one.

I mean spending less time on the mechanics of production and more time on the decisions surrounding it.

What is actually worth saying?

What evidence supports it?

What has changed?

Which audience cares?

What is the strongest version of the argument?

How does this connect to what the organisation already knows?

What should be developed further?

What should not be published at all?

Those decisions do not disappear when AI gets better. They become more important.

Because when producing another piece of content becomes almost frictionless, restraint becomes a competitive advantage.

Not an “AI-powered writer”

This is why I am increasingly sceptical of the idea that the future content professional is simply an AI-powered writer.

That framing keeps the job inside the production layer and just makes it faster.

The more interesting role looks closer to an editor, strategist, researcher and orchestrator combined. Someone who can work across a larger part of the problem precisely because the cost of executing each individual task has fallen.

Teams still matter. Specialist expertise matters. Different perspectives matter.

Great design, research, video, distribution, subject expertise and creative challenge do not suddenly become interchangeable, and anyone claiming otherwise has usually never had to do several of them properly.

But the unit of capability inside the team is changing.

One experienced person can increasingly do things that would previously have required much more infrastructure around them.

What this doesn’t settle

Two things follow that I cannot resolve, and I would rather name them than let the argument imply an answer it hasn’t earned.

The first is what any of this means for the size of a team.

The same observation supports two opposite conclusions. Fewer people producing what the organisation produces now. Or the same people producing work of a kind that was previously out of reach — deeper research, better evidence, more ambitious arguments, work that nobody had capacity to attempt.

AI does not decide which of those happens. Organisations do. And the choice tends to reveal what they thought content was for in the first place. An organisation that believed it was buying volume will take the saving. An organisation that believed it was buying judgement will buy more of it.

The second is that this argument rests entirely on the word “experienced”.

The capability expansion I am describing depends on somebody who already knows which evidence is weak, which argument is thin, which stakeholder request should be resisted and which brief is wrong. Give the same tools to somebody without that, and the result is more work produced faster with no improvement in the judgement of what deserves to exist.

So the expansion is not uniform. And it raises a question I do not think the industry has confronted: if the routine production work is what AI absorbs first, that work was also how people used to acquire the experience this model assumes.

The ladder that produced experienced operators is made of the rungs being removed.

I do not know what replaces it.

The better question

None of this tells organisations what to do with their structure.

It does suggest they are asking the wrong question.

Not:

How many content tasks can AI automate?

But:

What happens to the design of a content organisation when every talented person inside it becomes much more capable?

The first question produces a list of things to cut.

The second produces a much harder conversation about what the organisation is now able to attempt.

That is where I think the consequential change is happening.