The Rising Floor

A year ago, mediocre work still looked mediocre. Weak writing read as weak. Thin concepts looked thin. Generic interfaces looked unfinished. You could spot the gap between someone who had put in the reps and someone who hadn’t, often from across the room.

That gap is closing.

A single person with decent prompting can now produce work that looks polished in an afternoon. Design, writing, illustration, motion, code, strategy decks, product copy, research summaries, interfaces. The average output across every creative discipline is getting smoother and faster.

The floor is rising.

That changes the economics of creative work. It also changes something deeper about how we recognize expertise.

For most of my career as a designer, creative work operated on visible effort. Production quality was the currency. Senior people had taste, juniors handled execution volume, and the ladder depended on the scarcity of craft. You could see who was good because their artifacts were good.

That model is breaking.

When execution becomes abundant, the artifact stops telling you what it used to. A polished interface no longer proves product judgment. A finished deck no longer proves strategic clarity. Confident writing no longer proves clear thinking.

This is the part most senior creative people are still working through. The early instinct is to dismiss AI output as inherently shallow, to treat it as a junior approximation of the real thing. That belief falls apart the moment you spend serious time inside these systems. Some of the outputs are genuinely useful. Some are excellent. Some I would have been proud to ship five years ago.

The problem is not quality. The problem is discernment.

The real gap now lives between work that looks finished and work that actually solves the problem. That gap is invisible to anyone who only judges by the surface. That is most people, most of the time.

Coherence is not intelligence

I keep watching teams mistake coherence for intelligence in AI products. The interface feels smooth. The microcopy sounds confident. The interaction appears thoughtful. The empty states are charming. Then you use the product for five minutes and realize the system has no real grasp of what the user is trying to do.

The product is aesthetically complete and strategically hollow.

This is a failure mode that barely existed before. If a team had the craft to ship something this polished, they probably also had the discipline to figure out what it was for. Polish was expensive enough to filter out confusion.

A recent example. I sat through a demo of an AI writing assistant that did everything right at the surface level. The onboarding was beautiful. The empty state had a small, friendly illustration. The first prompt suggestion was crisp. By prompt three, the model was hallucinating internal team names that didn’t exist. The team had spent months on the visual system and weeks on the model behavior. You could feel the asymmetry the moment you used it.

That asymmetry is what design leadership is becoming. Not making the thing prettier. Catching the moment where the prettiness is doing rhetorical work the underlying system has not earned.

What happens to apprenticeship

The part I think about most is what this does to how people become senior.

For years, junior roles in creative fields were built around repetition. You laid out a hundred pages and started to see why some breathe and some don’t. You wrote a hundred headlines and started to feel which ones land. You shipped a hundred features and started to notice which ones cause the support queue to spike on Tuesday. Production taught pattern recognition. Pattern recognition built instinct. Instinct, over time, became taste.

The reps are now optional.

A young designer can generate fifty landing page variations before lunch. A junior writer can synthesize five positioning directions in ten minutes. A product team can mock up an entire flow without opening a design file. The friction that used to do the teaching has been removed.

This is the question I do not have a clean answer for. Taste still requires exposure. Judgment still requires friction. Good instincts still come from sitting with the discomfort of something not working and figuring out why. AI accelerates the output. It does not automatically accelerate the part of you that knows when the output is wrong.

I think we are about to live through a strange decade of people who can produce senior-looking work without ever having developed senior-level judgment, and the industry is going to spend a while figuring out the difference. The people who develop real discernment will do it on purpose, against the grain of how easy everything has become. They will take the friction the tools removed and add it back deliberately. Critiquing their own work harder, studying the things that did not work, and spending time with users instead of with finished artifacts.

I do not think anyone is currently designing apprenticeship programs around this. They probably should be.

Where seniority moves

If I had to name the defining senior skill of the next decade, it would not be craftsmanship. It would not be prompting. It would not be software fluency. It would be judgment.

The ability to look at ten plausible directions and recognize the one that actually works. The ability to hold human trust, narrative, pacing, ambiguity, interface behavior, emotional response, organizational reality, business incentives, and system constraints in your head at the same time when you are deciding what to ship.

That is harder than execution. It always was. Execution was just easier to measure.

The future creative organization, the one I want to work inside, is smaller and higher output, with fewer production bottlenecks. But heavier at the top, because the real value now sits with people who can shape intent before the machine generates the artifact. People who define quality up front instead of inspecting it on the way out. People who understand that the interface is no longer static, that system behavior is the product now, and that designing system behavior requires a different kind of authorship than designing screens.

The floor will keep rising. That is the easy prediction. The harder one is that the ceiling will rise too, but only for the people who treat judgment as a craft and practice it like one. Everyone else will produce competent, finished, hollow work and wonder why it isn’t landing.

For me, this is the most interesting moment my field has had in a long time. The tools got bigger. The job got smaller and more interesting at the same time. The thing that always mattered most, knowing what to make and why, is finally the thing that matters most visibly.

That is worth raising my own standards for.