But why taste?

Scroll through design X or LinkedIn and you'll see the same argument on repeat. AI can produce competent design. So what's left for designers? Taste. Taste is the moat. Taste is the thing AI cannot replicate, and so the future-fit designer is the one with the most of it.

The question worth asking isn't what humans still have over AI. It's the other way around. AI is not the complete answer. The work ahead is to embrace what these tools can do, and then look hard at what humans bring that the tools cannot.

To be clear, taste is a real skill. It takes years of absorbing creativity, refining, and developing a point of view. It's what separates a competent design from one that feels right for the brand it's serving. It's not that I think the taste discussion is wrong, it's just that I think it's incomplete.

I have this feeling that taste won't hold the line on its own. AI is getting better, and there's no good reason to think this is the one ridge it doesn't climb. Taste is also notoriously hard to measure. That pushes design closer to art, where judgement replaces metrics. But judgement at scale is a popularity contest. AI is built to win those, it averages, it aims for the 50th percentile, it gives you the version most people will nod at. That's not the design 50% think is the best or the most creative. It's the design the most people can agree with. No spike of emotion. No strong reaction in either direction. Just a quiet, broad "yeah, that's fine."

So if taste isn't the moat we thought it was, it begs the question. What does hold? I like how Magnifica Humanitas articulated it: machines lack human experiences, embodiment, and an understanding of love or responsibility. With just how much noise, excitement and disruption AI is making I think we need to reflect on what we have still to offer.

The first is philosophy. The habit of looking at a problem from every angle, reasoning from values, applying metaphors to help make sense of problems, asking why something matters and not just whether it works. Philosophy is what shapes the experience. It takes a problem that's too large or too tangled to see clearly and gives it a form. A metaphor, an analogy, the right framing, anything that lets other people see what you're seeing. That's how complex problems become ones a team can actually act on. And once the shape is there, the rest follows. Alignment when the data points two ways. Conviction when a stakeholder wants the obvious answer and the obvious answer is wrong. Philosophy is how you arrive at a point of view or make sense of complex systems. AI alone doesn't have a point of view. It has a distribution.

The second is judgement. AI gives you the 50th percentile by design. A designer's job, increasingly, is to know when to refuse it. When the model produces the version most people will nod at, judgement is the capacity to look at it, see what's missing, and choose the other way. Not for the sake of contrarianism. Because the consensus answer isn't always the right one, and the moments that matter most in design are usually the ones where it isn't. AI will zag every time, because the model is built to. Judgement is knowing when to zig.

Philosophy and judgement aren't enough on their own. The work is to make design calls that move a number, change a behaviour, win a customer. AI can write a strategy doc. It can't decide which design call is worth making, or tie that call to something the business will actually feel. That last part is the work I care about most.

None of this is an argument against AI. The opposite. AI is the most powerful tool a designer has ever had, and the designers who win the next decade are the ones who use it well. They'll use it to move faster, generate more options, test more ideas. But they'll bring their own philosophy, their own judgement, their own strategic intuition to decide which of those ideas is actually worth shipping.