Curation was always the job. AI just stripped the busy work.
The hype cycle around AI and creative work is still running hot, and it's running in one direction: replacement. AI will replace marketers. AI will replace designers. Nobody's talking about what's actually happening in the room.
What's actually happening is narrower than that, and it's on us to prove it. AI is stripping out the busy work, not the judgment. The people who survive this aren't the ones who resist it. They're the ones who use it to move faster and still show up with the thing AI can't produce on its own: the right call.
The busy work was never the value add
For most of my career, a huge share of the job was production: sketching, drafting, building variations, executing a direction once you'd already decided on it. That work took time, and because it did, it was mistaken for the value.
It wasn't. The value was always the decision before the production started — the direction, the tone, the narrative, the visual language. Knowing which idea to build in the first place. Production was just the tax you paid to find out if the decision was right. AI doesn't touch the decision. It removes the tax.
What's left is what was always the point
Take that tax away, and what's sitting there is the actual skill: picking the right direction out of several plausible ones, knowing which tone will land with this audience, recognizing the narrative that will actually move someone versus the one that just sounds good in the room.
That was never a lesser skill hiding behind production. It was the hard part the whole time. We just didn't have to prove it in isolation, because it came wrapped inside hours of execution that made it look like something else.
Generate headline ideas, layout compositions, and different campaign angles, and there's no execution left to hide behind. Just the decision. That's not a smaller job. It's the job with everything extra cut away. Numerous options. One right solution. That's the skill.
Why the decision is harder than it looks
A few things make this genuinely difficult, not just faster:
Competence hides the wrong answer. When AI generates options, most of them will be fine. Fine is the enemy here; it passes at first glance. The skill isn't spotting the bad ones; that’s the easy part. It's using your instinct to spot the right one at speed.
There's no brief for taste. You can brief a generation task on: audience, tone, constraints, and references. You can't brief "know it when you see it." That comes from years of pattern-matching against what's actually worked, what's actually landed, what's actually moved a customer or built a brand. AI has no access to that. It's the one input this work runs on that generation can't replace.
Volume tests conviction, not patience. The more options in front of you, the easier it is to talk yourself out of a strong first instinct. Knowing when to trust your read and stop looking is a harder call than it sounds when option forty-nine is sitting right there, promising it might be better.
Why this matters more than the hype cycle admits
Here's the part getting lost in all the "AI replaces the industry" noise: most clients don't actually want an AI-only solution. They want someone accountable for the call — someone who can look at the output and say, " This one, for this reason, and be right about it. Nobody's signing a contract with a model. They're signing one with the person who knows how to direct it.
That's the opening, and it's also the risk. If your value has been "I can produce things," AI just made that easier to replace than it's ever been. The people who treat AI as a shortcut to skip judgment (who ship the first fine option instead of doing the work to find the right one) are the ones who are actually vulnerable here. Not because AI replaces them directly, but because they've stopped doing the one thing that couldn't be automated in the first place.
The actual upside
AI isn't replacing the skill. It's giving it more room to work. Limited time and limited budget used to mean limited exploration; you could only afford to chase two or three directions before you had to commit. Now you can chase fifty and still land on the right one, because the cost of exploring is no longer the bottleneck. The judgment call is.
That's a real advantage for anyone whose value was always the decision, not the drafting. The market isn't short on raw materials anymore, and it never will be. What it's short on is people who can look at fifty options and consistently land on the one that's right for this brand, this audience, this moment.
That's not a fallback. That's the job, distilled down to the part that was always worth paying for, and the part you now have to prove, on purpose, instead of letting it hide inside the hours you used to spend making things.
AI can hand you numerous options all day long. It still can't tell you which one is right. That call is still yours, and it's the only one that was ever the job.
