A Nielsen Norman Group survey of 521 UX practitioners says the time AI saved rarely stayed with them: it was absorbed by higher output expectations or spent fixing other people’s AI-generated work. Figma’s randomized trial of Figma Make, by contrast, timed individuals on single tasks and found roughly 20% shorter completion times.
The episode walks through why a per-task stopwatch cannot register work that moves downstream, and the caveats: the survey’s comment-based findings come with a self-selected sample, and no source has measured net team time. The conclusion that per-task timing likely overstates team gains is an inference.