Skip to content
Header
All Episodes

A Pragmatic Approach to AI

Episode 12100:14:16 • 3.13.25

Kate Moran

VP of Research & Content, Nielsen Norman Group

Share
  • AI
  • Research

Kate Moran is VP of Research and Content at Nielsen Norman Group, where she's about to hit ten years. She got there the long way, through technical writing, content management, and front-end development, shadowing UX people at IBM and picking up a master's degree along the way. The thesis she was convinced no one would ever read is what got her recruited: Jakob Nielsen found it and wanted to publish it.

That research-first instinct shapes how her team has handled AI since 2023. NNG has stayed deliberately slow to comment on emerging patterns without evidence behind them, and Kate is blunt about what they found when they started testing the tools flooding the market. A lot of the claims are false. Research can't be replaced by asking a chatbot whether your site is good, and Figma is not dead. What generative AI does do is speed up a hundred small things in a day, which is a real gain and a very different claim from replacing a researcher.

On the UX of AI itself, Kate points to studies from Pew and Anthropic suggesting most of the American workforce still isn't using these tools day to day, and argues conversational interfaces are far less intuitive than designers assume. An empty text box hands the user all the work of figuring out what it's for. Her fix is narrower scope: Spotify's AI playlist generator can't write you a sonnet, and that constraint is exactly why people understand it on sight. For designers entering the field now, her advice is to stay flexible and not panic, since the confident predictions being made about AI and design jobs are educated guesses. Newcomers may even have an edge, because they're learning this once instead of unlearning a decade of habits.

Recommended Episodes

Want to be featured
on the pod?

Our guest list is still growing. If you're in design or product and have something you wish someone had told you earlier, we want to hear it.

Schedule an interview

No — it's changing what designers are paid for. AI has made producing screens and prototypes nearly free, but it can't decide whether a design is the right one, how a whole system should fit together, or what will actually move a user. That judgment is where design expertise now lives, and it's becoming more valuable, not less.

Less than the industry assumes. Studies from Pew and Anthropic suggest most of the American workforce still isn't using these tools day to day, which matters when you're designing for a general audience rather than for early adopters.

Stay flexible and don't panic. The confident predictions being made about AI and design jobs are educated guesses, not findings. Newcomers may even have an advantage: they're learning this once, rather than unlearning a decade of habits.

Pragmatically, and mostly in the hundred small places Kate describes. AI helps us move faster on first-pass synthesis, transcript cleanup, early prototypes, and exploring more directions than we could by hand. It doesn't decide what to study, interpret what findings mean, or judge whether a design is right — that stays with our researchers and designers. Every client engagement is still grounded in real users, and we're transparent with teams about where AI touched the work.