There is a moment I keep noticing lately.
Someone shares their screen, and within a minute there are twelve versions of the thing. Twelve homepage layouts. Twelve logo directions. Twelve subject lines for the same email. The energy in the room lifts, because it looks like progress. Then somebody asks, “Okay, so which one?”
And the room goes quiet.
That quiet is the whole story. We solved the wrong bottleneck.
Nobody said the work got better
A study came across my desk recently that put language to something I had been circling for months. A researcher ran eight group conversations with designers working in retail banking, government, consulting, tech, industrial design, academia, agencies, and independent practice, across Europe, the US, the Middle East, and Asia-Pacific. Different industries, different continents, different levels of seniority.
Not one participant said their work had gotten better.
Faster, yes. Prototyping collapsed from days to hours. Exploration got wider. But when asked directly whether the output improved, the answer was some version of what one visual design lead said: “It’s getting easier. But at the same time, I don’t think this is equal to better design.”
I want to sit with that for a second, because it runs against almost every promise being made in the market right now. The tools are delivering exactly what they advertised. They are producing more, sooner, cheaper. What they were never able to promise is the part that actually determines whether your business grows: whether the thing you made is the right thing.
The strange part about experience
Here is the finding that stopped me cold. An industrial designer in the study said, “the more you know what you want, the harder it is to work with it.”
Read that again if you have been quietly wondering why these tools feel clumsier to you than they seem to feel for everyone else.
The leverage runs backward from what we assumed. If your intent is vague, the tool fills the space and you feel like a genius. If your intent is specific, sharpened over years of watching what actually works with real customers, then every round of describing what you want is a round of compressing something precise into something approximate and handing it to a stranger to interpret.
One founder in the study named it plainly: “designers are not very good at communicating by language, and then you’re left with the interpretation of the language model.”
This is not a skill issue. It is a structural feature of the tools. Twenty years of pattern recognition does not fit in a prompt box.
From doing to describing
The deeper shift the research surfaced is that we moved from doing the work to describing the work. That sounds like a small change. It is not. Three things follow from it.
The first is loss of precision, which we just covered.
The second is quieter and more expensive. Craft has always been learned through making. You build the thing, you watch someone struggle with it, you understand something you could not have understood by thinking about it. When the making gets outsourced, that loop closes. The people entering our fields right now are getting output without getting the education that used to come attached to it. That bill arrives in about five years.
The third is the one I would flag hardest for anyone running a business. When thousands of people describe what they want using roughly the same vocabulary to roughly the same models, the results converge. Everything starts to look like everything else. Your competitor’s landing page and yours begin to rhyme, because you both asked for “clean, modern, trustworthy.”
Sameness is not a design problem. It is a revenue problem.
The time you saved is already gone
You might expect that all this speed bought people breathing room. It did not. The study found that the time saved was immediately absorbed by expanded scope and raised expectations. Deliver three concepts in the time you used to deliver one, and three concepts becomes the new baseline by Thursday.
So the bottleneck moved. It is no longer production. It is judgment. And judgment needs something that production never did, which is time to think.
One participant said, “I don’t have even time to reflect on what I’m thinking.”
The interfaces themselves work against reflection. There is always another suggestion, another variation, another regenerate button. There is no natural place to stop. Participants described having to build defenses on purpose: scheduling offline hours, declaring modes of work, closing the tools to hear their own thinking again.
What I would actually do with this
If speed is now abundant and discernment is now scarce, then the leaders who win are the ones who invest in discernment.
Practically, that looks like knowing your customer well enough to reject nine of the twelve options in under a minute. It looks like having an actual point of view about who you serve and what they need, so that convergence never touches you. It looks like protecting thinking time the way you protect revenue-generating time, because right now it is the same thing.
The people in that study were not afraid of being replaced. Not one of them said they could fold their hands and watch it work. What they described was subtler and more human: looking at their feet more than they used to, and never quite getting to the big moments anymore.
I do not think the answer is to use less of these tools. I think the answer is to be far more deliberate about what we are pointing them at, and to stop mistaking motion for progress.
So here is what I am curious about. Where have you felt this in your own work? Have you noticed yourself producing more and deciding less? And what have you done, if anything, to get your thinking time back?
I would like to hear it.
Study referenced: “How is AI Changing Design?” by Full Moon
Comments +