I've spent my career obsessing over UX—flows, friction, clarity, the moment a person either trusts you or bails.
Lately I've been thinking in MX: model experience.
Not instead of UX. Beneath it.
UX is still essential
If someone can't complete the task in front of them, nothing else matters.
Buttons. Hierarchy. Accessibility. Speed. The basics aren't basic—they're the contract.
But products aren't single screens anymore
They're loops. They're relationships. They're systems that remember, adapt, and sometimes guess.
When AI enters the product, you're not just designing interfaces. You're designing how the product thinks, explains itself, recovers from mistakes, and earns trust over time.
That's MX.
What MX includes
- How the model represents the user's intent
- How uncertainty is communicated (not hidden)
- How errors feel—human, fixable, not humiliating
- How memory works (what's stored, what's forgotten, what's visible)
- How the experience stays coherent across touchpoints
UX gets you through the door. MX decides whether people stay.
Why this matters for leaders
If your team still treats AI like a feature bolted onto an old flow, you'll ship demos—not products.
MX forces you to ask harder questions: What are we implying when the system answers? What happens when it's wrong? Who owns the relationship?
Those are product leadership questions. Not prompt engineering trivia.
Where I've felt this firsthand
Building Backline meant designing for musicians juggling chaos—schedules, people, last-minute changes, egos, gear, all of it.
A calendar UI wasn't enough. The model of the band, the gig, the commitment had to feel true. That's MX work dressed as UX work.
Takeaway
UX isn't dead. It's growing up.
The next leap is MX—designing the experience of the system itself, not just the chrome around it.
If you're a product leader still hiring for screens alone, you're already behind. Hire people who can think in models, not just mockups.
