Nearly nine in ten organizations now use AI in at least one business function. Ninety-four percent aren’t seeing significant value from it. Gale Robins, writing for UX Collective, argues that the gap is a framing problem, not an adoption problem. Her earlier piece on discovery judgment made the same case; the new one sharpens it with an anecdote that shows the trap:
A team I spoke with recently had compressed their discovery cycle from six weeks to ten days using AI. They were proud, and the throughput was real. When I asked what the work had taught them that they did not already believe, the answer was: not much. Same questions, faster. Same answers, sooner.
Same questions, faster. Same answers, sooner. Her analogy for the wider pattern is the electric factory one I’ve used before:
When factories first installed electricity, productivity barely moved. Manufacturers replaced steam engines with electric motors and kept the line-shaft layout. The breakthrough came later, when they redesigned the factory around what electricity made possible. The technology was only part of the answer.
Robins maps McKinsey’s three waves of AI value—productivity, differentiation, transaction-cost reduction—and finds most teams stuck in the first one. Robins on where they have to go to get out:
These decisions are upstream of every artifact a team produces. They are also where AI productivity gains help least, and where human judgment compounds the most.
Robins’s evidence undersells her own thesis. She leans on Generative AI at Work—the Stanford-and-MIT customer-support study by economists Erik Brynjolfsson, Danielle Li, and Lindsey Raymond that became the canonical citation for “AI helps novices most”—to argue AI raises the floor, not the ceiling. Novices gained 34%; experienced workers, basically zero. That’s why so many designers who have never coded—like me—are now suddenly shipping with this newfound superpower. It’s the same finding behind the junior designer crisis. But LinkedIn’s Full Stack Builder rollout found the opposite: top performers adopted AI fastest and got the most out of it, because they had the judgment to know what to ask for. The floor-not-ceiling story is only true where the questions are fixed. Once the questions are the work, the pattern inverts. That’s exactly the territory Robins is mapping. If AI rewards the experienced most when the work is judgment-shaped, framing is where the gap between teams widens.
