Product Design: From Creator to Curator

There's a brilliant episode of Lenny's Podcast with Noam Segal on how tech workers actually feel about AI right now. Unsurprisingly, the headline finding is that founder-executives are unfazed. The ESTPs of the world are busy adding another string to their bow.

Now, who is most affected?

Designers and researchers, of course.

The survey found people with careers in design and research were the most overwhelmed, tired and worried of everyone asked.

I am no exception. The more I use AI tools in my day-to-day, the more I find myself writing and sketching on paper, finding comfort in the freedom and familiarity of my own pen.

So I’m here writing this article as a way to try and make sense of it all. I have a hunch that something about the way designers think is our most valuable asset in this changing world.

There's little I find more relieving than the hushed silence of an art gallery. Hours spent drifting through sparse white rooms, using the artist's window to see the world through a new lens.

But what is art, actually?

Let's be honest, beauty itself isn’t always the priority. Anyone who's stood in front of a Tracey Emin knows that with certainty. Perhaps you could argue it's about finding the beauty in things? If I were to sum it up, I’d say art is about connection. Using something that appeals to the senses to represent the human experience in a new and interesting way, to elicit an emotion.

Machines producing art is a philosophical debate that deserves its own multi-page essay, so please give me the benefit of the doubt as I heavily simplify. Right now my opinion sits somewhere here: the continuous system of the human mind, boundless emotion, suffering, empathy, and wonder feels essential to creating true art. I'm not sure how deep you can go when the creator is constrained to a world of 1s and 0s.

Product design, of course, is different. We want the interface to have elegance, but the real work for the past 30 years has been adapting that interface to how people actually use it. We built entire industries around bending a rigid UI to the messiness of human behaviour. UX, HCI, IxD exist primarily for this reason.

Today, AI knows all about heuristics and best practices; it knows the context and behaviour of your users, it has ingested all historical research, and it can both read and write to your codebase. Seemingly overnight, a simple prompt of one or two lines can generate an entire app, and let's be honest, the output is actually pretty good.

Some will say, “But everything will look the same! What about creativity?” Unfortunately for us designers, AI can be good at that too, if you know how to unlock it. One of my favourite prompts is "give me answers of varying probabilities." There are countless articles on this; one of my favourites is here (Lenny's again).

No wonder we're all anxious, depressed, destabilised.

However, the question I keep coming back to is this: where is the art in product design? Where is it essential that we tap into the power of our complex, continuous human experience? Where are there gaps that need to be filled with the boundless imagination of a designer's mind?

So I went looking for something firmer than a feeling.

Back in the 1980s, a study took place to assess whether people trained in design approached problems differently from those with a background in science. Design students and science students were given the same puzzle: arrange coloured blocks to satisfy a hidden rule. The scientists tested combinations systematically to work out the rule first. The designers did the opposite: they proposed an arrangement, saw if it worked, adjusted, and kept going until one succeeded, learning the rule almost by accident along the way. The conclusion was clear: scientists solved by analysis. Designers solved by synthesis.

No one has run this exact experiment on AI (yet), but comparable hidden-rule studies show that many AI models behave like the scientists. They search systematically, testing each hypothesis and using the findings to pick the next move. Interestingly, some models without structured reasoning did something else entirely: they produced a plausible-looking answer based on patterns they had seen before, without reliably tracking what had already been ruled out, failing to eliminate the wrong hypotheses. A weakness that has been widely documented.

Crucially, none of the models approached problems the same way as designers. Designers learn about the nature of a problem by trying out solutions, not by setting out to study the problem itself. They propose something, then refine until it works, using their own judgement.

There's no question design is changing. With AI, creating anything from scratch is yesterday's process. Some new design tools are even stripping out direct manipulation entirely.

However, convenience comes with a catch: in an industry where creating five solutions takes five minutes, a world of endless options quickly becomes a trap, making it easier than ever to walk confidently in the wrong direction. If art is about transforming experience into connection, then judgement (choosing well among endless options), is its most practical form.

The designer mindset is built for this. The ability to work through an evolving problem space, turning the abstract into something both concrete and visual, then refining it using judgement built from years of training and experience. Tapping into this doesn't just lead us in the right direction, it carries us past the edge of what's known. It's exactly what most real-world business problems ask for. They don't come with a clean, fixed set of rules; by nature, they're ambiguous and ever-shifting. I'd argue the same skill set matters far beyond product and is just as valuable against wider strategic and organisational problems.

So, let's celebrate the messiness of our human-ness. In a world where creation is now free, our judgement becomes the art.