Why is UX research more important in the AI era?

Building is no longer the bottleneck. Anyone can ship fast. The bottleneck is knowing what to build. AI generalizes well and designs for the median user, while outliers often hold the insight. Our job is to find those deviations.

How do you spot when research leads to the wrong conclusions?

Warning signs: method chosen before the question is clear, wrong or tiny sample, leading questions, yes/no-only answers. Best guardrails: hypotheses written upfront, multiple reviewers in analysis, and qualitative work alongside surveys.

What should AI do in UX research, and what should it not?

AI is strong at transcription, competitor scanning, and initial theming. Decisions from research findings must still be made by people. Give AI the repetitive work; do not outsource judgment.

Back to episodes

Is UX Research Dead in the AI Era?

3 March 2026

This episode is in Finnish. The video below is the original recording; this page provides an English summary for easier reading.

Listen on Spotify

Summary

Is UX research obsolete when AI builds fast? Tuomas and Annika on data, humans, AI limits, and why research matters more, not less.

About The Good Side in this episode

Tuomas Kaartoluoma co-founded The Good Side with Mikki after meeting at Haltu. He brought strong UX research capability, now a core strength of the studio, alongside long-term clients who moved with them from his earlier practice.

Key discussion points

UX research matters more in the AI era: anyone can build; the bottleneck is knowing what to build
AI designs for the median user; the most valuable findings often come from outliers AI misses
Wrong research conclusions create a snowball: strategy accelerates in the wrong direction
Good researchers work implicitly; AI works explicitly: interpreting emotion and reaction still needs humans
Say-do gap: what people do, what they say, and what they say they do are three different things
AI fits transcription and first-pass theming; conclusions from research must still be human-made
Jakob’s Law: users spend most time in other products; benchmark there, not only against direct competitors

Questions & answers

Why is UX research more important in the AI era?: Building is no longer the bottleneck. Anyone can ship fast. The bottleneck is knowing what to build. AI generalizes well and designs for the median user, while outliers often hold the insight. Our job is to find those deviations.
How do you spot when research leads to the wrong conclusions?: Warning signs: method chosen before the question is clear, wrong or tiny sample, leading questions, yes/no-only answers. Best guardrails: hypotheses written upfront, multiple reviewers in analysis, and qualitative work alongside surveys.
What should AI do in UX research, and what should it not?: AI is strong at transcription, competitor scanning, and initial theming. Decisions from research findings must still be made by people. Give AI the repetitive work; do not outsource judgment.

Topics: UX research, user research, AI and UX, user interviews, research methods, say-do gap, usability testing, The Good Side, Behind the Pixels, research plan

Is UX Research Dead in the AI Era? | Pikselien Takana | The Good Side