AI didn't make design less important. It made it more important — and fundamentally different.​

When anyone can generate a product, the question shifts from ‘can we build it’ to ‘should we build it.’ We go from calculating the volume of output to the value of outcomes. AI can tell you what users click, but it can't tell you why they abandon a flow. And even more critically, AI can generate an interface, but can't tell you whether it solves the right problem.​

AI can create dozens of potential directions in an instant. Determining which path will produce the outcome you want requires human judgement to pull together the elements of brand strategy, context, understanding of what matters to your customer, the competitive landscape, and an eye to what comes next. Expert prompt engineering doesn’t guarantee a worthwhile result. ​

The teams we work with who are seeing the ROI from AI enabled design processes aren't the ones moving fastest. They are the ones who slowed down at the right moments — to research the opportunity, question assumptions, design systems that are worth building, and test their own intuition with real people. That rigor is what makes AI-accelerated processes actually deliver something worth having. ​

Some say “build it and see what happens,” but hope is not a strategy. This approach devours tokens while missing the mark with customers and users, affecting market position, brand perception, and ultimately customer loyalty. ​

What's one assumption your team is making about your users that you haven't validated recently?​