Future of User Interviews in AI Era
During one of the usability sessions for a product I’m building (Supalink — Your Identity Builder), the very first user started running into a-lot of usability issues.
They were not able to understand how to use drag & drop interaction, from where to preview their edits and many more.
Those flows that made perfect sense in my head were not working for them at all 🥲
In era before AI,
I’d take notes. Highlight the issues. Thank them for their time. Promise that the fixes would be part of the next release. And then… we’d both move on.
But this time was different.
I opened my Cursor AI editor to make the required changes as per the feedbacks, pushed the changes on the usability branch.
A few minutes later, I asked them to refresh.
Their feedback was already live. The expression on their face changed immediately.
It was not an interview anymore.
Seeing the feedback LIVE, The dynamic completely shifted.
They no longer felt like a participant helping me improve my product. They suddenly felt like they were building it with me.
Once they realized their suggestions weren’t disappearing into a backlog somewhere, they became dramatically more invested.
They started suggesting me what would improve my product dramatically, things I already had in my plan to release later but users needed them Right Now & that information is GOLD.
Instead of giving feedback, they started co-designing.
AI didn’t replace the interview. It supercharged it.
There’s a lot of discussion about AI replacing designers, developers or product managers.
I think we’re asking the wrong question.
Get Anupam Pareek’s stories in your inbox
Join Medium for free to get updates from this writer.
The real opportunity isn’t replacing the people involved in product creation. It’s reducing the delay between thinking, building, learning and improving.
For years, the loop looked something like this:
Feedback → Documentation → Prioritization → Development → QA → Release → User tests again
Sometimes that cycle takes weeks. Sometimes months.
Now imagine this instead:
Feedback → Build → Refresh → Continue the conversation
The interview doesn’t end because you discovered a problem. It evolves because you’ve already solved it.
Honestly it made me rethink on how anyone should be doing User Research
Traditionally, user interviews have been about observation.
Watch. Listen. Take notes. Don’t interrupt. Go back and analyse later.
That process still has value.
But AI introduces something entirely new. For certain stages of product development, interviews can become collaborative working sessions rather than passive observation sessions.
Almost like pair programming.
The user explains the problem. You explore solutions together.
The product evolves while the conversation is still happening. That’s a completely different experience.
The future feels closer than I expected
People often talk about AI making development faster. That’s true.
But I think the bigger shift is something else.
It’s shrinking the gap between what users expect and how the product actually behaves.
The shorter that distance becomes, the more human; product actually feels.
Because instead of collecting feedback to revisit later, I’m starting to build with people rather than merely for them.
