Legit question: When someone asks for “Design Experience with AI,” what are they actually asking for?
Generative chat interfaces create the illusion of design contribution at dangerous scale. In seconds, “designs” appear—on-brand colors, tortuous user flows—rushing onto the product roadmap like a best of compilation of Russian dash cam footage. What they skip is the hard (fun?) part: discovering which assumptions are wrong, figuring out what the user is really trying to accomplish with that Post-it note stuck to their monitor, and—most importantly—identifying the actual business problem.
Design is a process, not a deliverable. I’m far from the first person to say that. So why is this still up for debate?
I’ve been focused on improving human communication and shared understanding since the days of Microsoft Publisher. I watched Quark fall. I watched in horror as all the tools mimicked InVision patterns for interaction design (somebody must have owned stock in overlays). I’ve begged my Figma rep for table tool as robust as InDesign had 15 years ago. I’ve mourned as Axure ignored scalable style support.
I’ve built small apps for reserving hunting parcels during Dove season. I’ve helped scale remote health services tools that led to an acquisition for a truly silly number. This may sound like a weak humblebrag, but it’s context: I’ve been doing this for a long time, across wildly different problems/scales.
Affinity bias may frustrate retail users accustomed to the old way of doing things—but that’s a whisper compared to the gnashing of teeth that filled the halls when you ask someone to do their job differently. That’s high-stakes design advocacy. That’s the VIP room…you must be THIS 🙋 empathetic to ride the ride.
The design process creates psychological safety around those shifts. It gives executives visibility into what’s actually happening—and what meaningful change will really cost and unlock. In the end, this is a people problem. But technology is a convenient distraction. It’s emotionally easier to blame a neutral third party (“that stupid app”) than to admit, “Turns out we’ve been doing this the dumb way.”
I’ve used Chatty-G to free data from PDF prisons. She’s been helpful comparing product specs so I can decide I don’t actually need a robot vacuum. She is a powerful tool for retrieving concise information from the hoarder’s basement that is the sponsored-content internet. It feels like what Google search should be if advertising revenue weren’t strategically lowering the value of returned results.
Computers are excellent at computing. They are nowhere close to navigating the messy, political, emotionally loaded complexity of human culture—especially the version we inhabit at work.
So to close this loop: when people say they want “AI product design expertise,” what do they really mean?
Beyond using it as a powerful API to populate the presentation layer…what are they actually asking for?