Conversations Still Happen at the Speed of Humans
AI can build UI faster than I ever could by hand. It can turn rough interactive flows from a paragraph of text and explore multiple layout directions in minutes. That production power is real and I won’t pretend otherwise. But none of it matters if you haven’t had the right conversations first. Conversations with users still happen at the speed of humans. No model can replace the nuance you pick up when someone explains — slowly, sometimes poorly — why a product frustrates them. AI can help you do more with what you’ve learned, but it cannot replace the learning itself. The sequence still matters:
Understand the problem
Understand the user
Understand the business value you’re trying to create
Bring AI in to explore solutions
Skip that order and you get solutions in search of problems, just shipped sooner.
Research Faster with Clarity
Design has always included a quiet tax: synthesis. Synthesis is the mental work of making sense of scattered research and turning raw observations, notes, and hunches into clear, actionable insights. Research includes interview notes open across six tabs, session observations that feel meaningful but won’t quite resolve into a theme and half-formed patterns you sense but can’t yet name.
AI rarely surfaced something I hadn’t noticed at all. But it regularly helped me articulate what I already sensed. That naming mattered as it moved me from “I have a hunch” to “here’s a framed insight we can pressure-test.” The real gain in leveraging AI in design practices wasn’t time saved. It was confidence earned through clearer thinking.
Ideation: Widening the Space Before Narrowing it
Real constraints compress divergence. You sketch four directions instead of fourteen, and refine the selected one before you’ve truly explored. AI changed how I handle that compression.
I can now generate a wide spread of directions quickly, not as answers, but as provocations. Some designs are obvious, some are wrong, and a few are genuinely surprising. Those surprising ones rarely survive intact, but they seed directions I wouldn’t have reached on my own in the time I had.
The human stays in the loop as the decision maker. What AI produces is raw material. What you choose to pursue is the design.
Prototyping and Critique: Learning Sooner
AI can make rough prototypes from written descriptions, sometimes within hours of a framing conversation. Not pixel-perfect but real enough. Their purpose is to create that “30-second wow”: something tangible and believable enough to provoke genuine reactions from users early in the process. Those reactions move work forward faster than internal debate ever will. The speed isn’t the point. Learning from users earlier is.
Critique works the same way. As I generate and iterate with AI, critique is embedded in the back-and-forth, pushing on assumptions, asking for alternatives, and challenging the reasoning behind a proposed direction. That exchange isn’t about AI having the right answer. It’s about forcing clearer articulation of intent, tradeoffs, and risk.
By the time work reaches a team review, it’s already been pressure-tested. The conversation shifts from “what is this?” to “is this the right call?”, which is where human judgment and collective experience matter most.
What I’m Still Learning
I want to be direct: I’m still early. The skills required to work well with AI are changing fast, and my current ability — especially with agentic workflows — is nowhere near where it will need to be.
Agentic AI introduces design questions I don’t have clean answers to yet: How do you design for processes that partially automate themselves? How do you keep humans meaningfully in the loop when the loop moves faster? I’m working toward those answers and I think admitting that matters more than pretending I’ve arrived.
The designers who will navigate these questions and situations would be the ones who stay curious, stay honest about what they don’t know, and keep users, not technology, at the center of every decision.
“What AI produces is raw material. What you choose to pursue is the design.”
AI is shifting the boundary of what requires a human, faster than I expected. The job isn’t to resist that change. It’s to stay clear about what the change is actually for: better outcomes for the people we’re designing for, not just shorter timelines on a project plan.
And to stay honest with ourselves and our teams about how much there still is to learn. I’d love to hear how others are approaching this, what’s working, what isn’t, and what questions you’re still sitting with.
AI won't replace the conversations, the synthesis, or the judgment calls that make design meaningful. But it will keep raising the bar for how designers think, explore, and articulate their decisions. Want to build AI capability into your design & development practice without losing what makes the work worth doing? Let's talk about what that looks like for your team.
