Human Curation in the Age of AI: Scarcity Has Changed Locations

For decades, the advantage belonged to those who produced. Those who could write more, draw more, publish more, took up more space and captured more value. Production capacity was the bottleneck, and whoever unlocked the bottleneck won. Artificial intelligence undid this game in one pass.

Producing has become a commodity. Everyone can now generate, in any volume, with decent technical quality. And the economic rule is relentless: when something becomes abundant, value migrates to what has become scarce around it. What became scarce is not production. It's the trial. Curation, good taste and context became the new bottleneck, and therefore the new value.

Scarcity has changed places

It is worth understanding the mechanism, because it reorganizes priorities. In a world of production scarcity, valuable work was to do. In a world of abundant production, the valuable work is choosing. When there are a thousand options generated in seconds, the bottleneck is not generating the thousand-first: it is knowing which of the thousand is good.

This is an inversion of hierarchy. The skill that previously seemed secondary, deciding what deserves to exist, deciding what to cut, deciding what goes in front of the public, has become the central skill. Editing started to be worth more than writing. Select, more than produce. The red pen became more powerful than the typewriter.

For leaders, the consequence is direct. The valuable professional of the next decade is not the one who delivers the most output, because output has become cheap. It's the one who has the judgment to say no, to point out what matters, to protect the public from excess. Hiring for volume is betting on the commodity. Hiring for discernment is betting on what is scarce.

Good taste is not luxury, it is function

There is an old prejudice of treating "good taste" as something decorative, subjective, without rigor. This prejudice became expensive. Good taste is the trained ability to distinguish what is good from what just looks good, and in a flood of plausible content this ability is what separates signal from noise.

AI excels at producing the plausible. She gets the average right, the expected form, the correct result enough. But the average plausible is precisely what doesn't stand out. Distinguish, within the sea of correct things, the one that is right for this public, at this moment, with this intention: that is taste, and the machine doesn't have that.

Taste doesn't fall from the sky, it's built. It is the result of exposure, comparison, error, accumulated reference. Anyone who has seen a lot of good things and a lot of bad things develops a filter that works faster than conscious explanation. This filter is what counts, and it is deeply human because it depends on a lifetime of particular experiences.

Context is what the machine doesn't have

AI processes the whole world and does not know yours. She doesn't know what happened in the last meeting, what her client's sensitivity is, what has already been tried and failed, which joke goes down well with this audience and which offends. Context is situated information, and the machine operates on general information.

Quality curation is, in essence, application of context. It's taking a technically correct piece and asking: correct for who, for when, for what objective. The same phrase is brilliant in one place and disastrous in another, and only those who know the terrain know the difference. This knowledge is not in any dataset, it is in the head of whoever experienced the situation.

This is why the irreplaceable curator is the one with depth of context. The editor who knows the reader. The leader who knows the team. The creator who knows his community to the point of knowing what it will love and what it will reject. This someone delivers something that no scale can solve: situated relevance. I address exhaustion with content without context in anti-AI slop and the value of the human.

Responsibility: someone has to sign

There is a dimension of curation that is often overlooked and that is perhaps the most important: responsibility. AI generates but does not respond. She has no reputation to lose, no consequences, no accountability. Someone needs to put their name at the bottom and say “this is going because I decided it is going to”.

This signature is what gives weight to a choice. When a human heals, he assumes the risk of judgment: if it is bad, his credibility falls. This commitment is exactly what is missing in mass-generated content, which has no owner or culprit. Responsibility is the counterpart of judgment, and it is irreducibly human.

For those who lead, this redefines the role. Leading in the age of AI is, in large part, about being the responsible curator: approving, vetting, prioritizing, and answering for the whole. The machine expands the capacity to produce options; the leader provides the discernment to choose and the courage to sign. Whoever outsources the judgment to the tool gives up precisely the part that cannot be outsourced.

What this means for creators and leaders

For creators, the message is liberating. You don't need to beat AI in volume, and you won't. You need to win in selection: what you choose to show, what you decide to cut, the angle that only you would see. The creator of the future is less a content factory and more a curator with a voice, someone the public trusts to filter the world.

For leaders, reorganization is a value criterion. Stop measuring the team by quantity delivered and start measuring it by quality of decisions. Reward those who cut out the superfluous, those who point out the error that no one saw, those who have the courage to say "that's not good enough" in the face of something technically correct. This is the talent that has become rare.

The turn I defend in this essay dialogues with the idea that human presence has become a luxury, which I developed in the new luxury is to appear human. They are two sides of the same coin: when producing is free, humans are worth what they decide, not what they manufacture.

Curation, taste, context, and accountability are not nostalgic skills that stubbornly survive AI. These are the skills that AI makes most valuable, because it solves everything around them and leaves exactly those ones uncovered. In a world where producing no longer impresses anyone, investing in judgment is what differentiates those who are sought after from those who are ignored.