I Was Wrong About This in 2015: The Shrinking Territory of Human Creativity

In May 2015, I wrote an article called Human Data. The subtitle was "a very optimistic vision of why humans are still needed." I was confident. I had just come back from SXSW, I was surrounded by the energy of the industry, and I believed — genuinely — that the depth of human intuition was something no algorithm could replicate. That qualitative research, ethnographic fieldwork, the ability to read a facial expression or the tone behind a word, would always give us an edge.

I was wrong. Or at least, I was more right than I deserved to be, for reasons I didn't fully understand at the time.

What I argued then

The core of that article was simple: data tells you what, but not why. Analytics can tell you that someone clicked on the red button instead of the green one, but it can't tell you what red made them feel, or what memory it triggered, or why that feeling translated into a purchase. I argued that the advertising industry was becoming too dependent on quantitative measurement and losing the ability to ask deeper questions — the kind you can only answer by sitting with people, observing them, listening to them over time.

That argument still holds. But it's no longer the most important question.

What I see today

I use AI every day. Not as a replacement for my judgment, but as something close to a very capable assistant — one that drafts, suggests, resolves, and produces at a speed that would have seemed implausible a decade ago. Technical questions that would have sent me down a rabbit hole of tutorials now get answered in seconds. Editing decisions that used to require a specialist are within reach. The output is genuinely better, and the time saved is real.

What I'm describing right now — this article — is part of that process. And that's exactly where it gets uncomfortable.

Because here is what I've noticed: the check is still mine. The judgment of whether something makes sense, whether it sounds like me, whether it serves the right purpose — that still requires a human in the loop. For now.

But I can see the horizon. And the horizon is closer than I thought.

The thing nobody is saying out loud

There's an idea that's been forming in my head that I haven't seen articulated clearly anywhere, so I'll try here:

When AI makes it possible for anyone to produce a near-perfect output, the average quality of everything goes up. A small business owner with no design background can now produce visuals that would have required a professional agency five years ago. A non-native speaker can write with fluency that would have taken years to develop. A junior publicist can produce a strategic brief that reads like it came from someone with a decade of experience.

This is remarkable. And it is also destabilizing.

If everyone's output reaches the same level of quality, differentiation disappears — or at least, it moves. The question stops being can you produce something good? and becomes can you produce something that couldn't have come from anyone else? The territory of human value shrinks to the things that are genuinely irreplaceable: taste, context, lived experience, the ability to know when something is technically correct but fundamentally wrong.

That's a much smaller territory than I was defending in 2015.

What might be left

I want to be honest: I don't have a clean answer. I'm not going to tell you that creativity is uniquely human and always will be, because I'm not sure that's true anymore. I'm not going to tell you that AI will replace everything, because I don't believe that either — at least not yet, and not in the ways that matter most.

What I do believe is that the people who will navigate this transition best are not the ones who resist AI, nor the ones who surrender to it entirely. They are the ones who understand what they are actually contributing when they put their name on something — and who are honest enough to keep asking that question as the answer keeps changing.

In advertising, we spent years talking about the importance of authentic storytelling. Of understanding the real human being behind the data point. Of asking not just what people do, but why.

That instinct wasn't wrong. It just didn't anticipate that one day the entity generating the stories might not be human at all.

The question I'm sitting with

When everyone can produce a perfect output, what are you, specifically, worth?

I don't think that's a question with a permanent answer. I think it's a question you have to keep earning the right to answer — through the quality of your judgment, the specificity of your perspective, and the honesty to admit when you were wrong.

I was wrong in 2015. I'm probably wrong about something in this article too.

The difference is that this time, I'm not sure I'll have ten years to find out.