AI Still Needs a Gentle Nudge—And That’s Good News for UX
Why AI’s inability to 'think laterally' is actually the best news for the future of user experience design
On my dog walk last night, I found myself in a back-and-forth with Gemini about diversity, innovation, and whether companies actually perform better when their leadership teams aren’t carbon copies of one another. The data was encouraging: Boston Consulting Group found that diverse leadership teams earn meaningfully more revenue from new products, and research suggests that educational diversity, in particular, sparks more novel work.
Different perspectives in a room, it turns out, produce better thinking. (Shocking, I know.)
I asked Gemini to drop the study citations into a Google Doc. It told me it couldn’t; apparently, it lacked doc-creation powers. Fine. But I was walking dogs and holding no pen, so I pushed: "Surely there’s another way to get this information to me?" It didn’t offer one. It just volunteered to recite the titles so I could Google them later.
A minute later, I remembered Gemini had scheduled calendar events for me before. “Why don’t you put the citations in a calendar invite for tomorrow morning?”
"Perfect," it said. Done.
So I asked the obvious question: Why didn’t you come up with that?
Its answer was honest and revealing. It admitted it had been narrowly focused on the "Google Doc" request and failed to think laterally across its own toolkit. It needed, in its words, “a gentle nudge.” It admitted that truly novel connections—the kind that jump categories from document to calendar—are still hard for AI. Humans remain the masters of adaptive, lateral thinking.
I work in User Experience design, where many colleagues are nervous that AI is coming for their jobs. But the same conversation that praised diverse teams for innovation also demonstrated why a single model struggles to produce that kind of lateral leap. Diversity of perspective is a feature of humans in a room together; it isn’t a property any one model has yet.
The best partnership is exactly what happened on my walk: the AI generating options quickly, and the human choosing, refining, and adding context. In UX, that context is empathy. It’s understanding a user’s frustration, their life outside the screen, and the unspoken reason they abandoned a form on step three. AI can speed the work up, but it can’t feel the user.
So keep using these tools. Keep giving them the nudge. Your judgment is still the most valuable thing in the room.