UX Collective: Never mind the prompts, here's the thinking
Dan Maccarone, founder of a design studio, spent a year rebuilding his team’s entire design process around AI. The result was not faster output. That, he writes in this UX Collective piece, turned out to be the right outcome — and the wrong expectation to have started with.
The article is a short, practical reflection on what actually changes when a design studio adopts AI seriously, rather than using it as a drafting shortcut. It challenges the framing most teams bring to AI adoption: that the primary return is speed.
The core reframe
Maccarone argues that the relevant question is not “how fast can AI make us?” but “what do we do with the time it provides?” The two framings lead to different workflows. Speed-focused adoption treats AI as a tool for completing tasks faster. The reframe he describes treats AI as a way to reclaim time for the parts of design that AI cannot do: holding a considered point of view, deciding what to make before making it, and knowing when a direction is wrong.
The workflow he describes
His studio’s approach has reasoning go in before the first prompt is written. This means the design rationale — what problem is being solved, what constraints apply, what the output needs to accomplish — exists as an artifact before any AI interaction begins. Once a prototype exists, the reasoning comes back out: the team uses it to evaluate what the AI produced against what they actually intended. Each time the prototype changes, the reasoning is revisited.
The practical effect is that the team never loses track of their own thinking during rapid iteration. When AI generates a variation quickly, there is a written basis for deciding whether it is better or just different. This prevents a pattern that Maccarone names but many teams experience without naming: knowledge transfer loss, where the accumulated understanding of why decisions were made evaporates as work moves fast.
What it does not cover
The article does not provide a template or specific tooling recommendations. It describes a mindset and a workflow pattern, not a step-by-step process. Teams looking for prompt libraries or AI design tool comparisons will not find them here.
Who it is useful for
Useful for design leads and studio founders deciding how to position AI within their team’s workflow, particularly those who have tried AI adoption and found it disorienting rather than clearly beneficial. Also useful for designers who sense that working faster is not the same as working better, and want a frame for thinking about why.