Better Than Nothing: Why AI-Led UX Research Deserves a Seat at the Table
- bradyux
- May 3
- 2 min read
by Brady Starr, Brady UX

Let’s get this out of the way: AI is not a substitute for real, messy, human conversations.It doesn’t sigh. It doesn’t hesitate. It won’t tell you your onboarding flow “feels weird.”
But you know what else doesn’t do any of that?The UX research most startups aren’t doing at all.
Right now, we’re living in a bizarre reality where companies will ship entire products based on vibes, gut feelings, or whatever the loudest stakeholder said in a Slack thread at 2AM — while simultaneously rejecting AI tools because they’re “not authentic enough.”
Let me say this clearly: AI-powered research is better than nothing. And “nothing” is exactly what many teams are doing in the name of speed or budget.
The False Binary
Some folks like to frame the conversation like this:
“Would you rather have AI insights or real user interviews?”
Yes, obviously I’d take real user interviews — any day. But that’s not the actual choice on the table for most teams.
The real choice is:
“Would you rather use AI to simulate a few user reactions based on tons of training data… or base your product direction on your own guesswork, or worse, no direction at all?”
Suddenly, that AI tool doesn’t sound so bad, does it?
A Starting Point with Superpowers
When we use AI at Brady UX, it’s not to replace human-centered design — it’s to accelerate it.
We use it to:
Draft proto-personas before we’ve scheduled interviews
Run predictive usability tests on early designs
Simulate user flows and friction points
Generate questions we might not have thought to ask
Catch biases baked into our own assumptions
No tool replaces judgment. But AI can amplify it — by giving us a richer head start. And in some cases? It actually unlocks paths to UX that might be more delightful than what we would've come up with on our own.
Unexpected Delight, Engineered by Machines?
Here’s where it gets exciting.
AI doesn't carry your team’s blind spots. It doesn’t know your industry jargon or that Karen from legal hates purple buttons. It’s trained on billions of interactions — meaning sometimes it suggests a pattern or path you never would’ve considered.
We’ve had moments where an AI-generated UX suggestion made us pause and go, “Wait… that’s kind of brilliant.” Not perfect. Not polished. But the seed of something new.
Final Thought: AI Isn't Cheating — Apathy Is
We’re not here to preach AI as a panacea. If you have access to real users, talk to them. Watch them struggle. Learn from their pain. That’s the gold.
But if you’re not doing research because “there’s no time” or “we’ll figure it out post-MVP”? Then you’re gambling with your product’s success.
AI isn’t cheating. It’s a shortcut through the fog — a way to move smarter when you don’t have all the pieces yet. And right now, that’s a hell of a lot better than standing still.
Want to test AI-powered UX research on your product?Let’s talk: bradyux.com/meet
Comments