Vibe Coding in UX Design or an Inevitable Evolution?

Vibe coding in UX design is a concept that raises more questions than it seems. Recently, talking with my brother, he mentioned that what I’ve been doing with AI while designing **Classic HEX** could be called that. So I went to read what it actually meant… and I didn’t like it at all.

What vibe coding in UX design is

The simplified definition was something like “coding without knowing how to code.” And that’s where I felt resistance. Because that’s not my case. I do know how to code. I just spend much more time thinking than writing code now.

So the question shifted: if it’s not exactly that… what is actually happening?

Because what I feel is not superficiality or improvisation. Quite the opposite. I feel more inside the process than before, just from a different place. And there’s a moment when it becomes obvious: things have changed. Again.

Will AI take my job? As Don Norman wrote on LinkedIn:

“Will AI take away my job? I always say, if you keep trying to do the same job? Yes, it will. But if you learn to use AI properly, you will be even more in demand.”

So… the work has shifted

Before, a large part of the work was execution. Writing code, adjusting details, solving problems step by step. There was friction, time invested, and that made the effort tangible.

Now the flow is different. You think of a solution, describe it, generate a version, review it, and adjust it. The process is faster, but also more mental. The effort doesn’t disappear — it just moves somewhere else.

It’s no longer about doing as much as it is about deciding.

And deciding well requires judgment. Understanding the system, anticipating behaviors, and knowing when something makes sense and when it doesn’t.

A more fluid (and lighter) workflow

The tools have also changed. Designers have tools like Figma Make or Google Stitch. Developers work with Copilot, Codex, Claude… Everything happens faster, with less friction.

In my experience, AI doesn’t just save time — it changes how I manage attention. I can move between tasks more easily, without that constant feeling of overload that used to come with multitasking. The flow feels more continuous, lighter.

And in that sense, more productive. But there’s something curious. The result is good, and still, the process doesn’t feel the same.

What shifts internally

New doubts start to appear — things that weren’t so present before.

• Effort = value We still associate value with effort. If something comes out quickly, it creates distrust, even if it works better.
• Blurred authorship Even if you make all the decisions, part of the result isn’t built directly by you, and that creates some distance.
• Loss of control You don’t see every step or understand every line. The system works, but it’s not fully transparent.
• Impostor 2.0 It’s not just “am I good enough?”, but “does this even count as real work?” A subtler doubt, but quite common.

Guilt (and where it comes from)

In the middle of all this, an unexpected feeling appears: guilt. As if using these tools well was, in some way, cheating.

Here I go back again to Don Norman. In design, he insists that we shouldn’t blame the user when something doesn’t work, but instead look at the system.

“Good design means the technology has been designed to fit people, not for people to have to learn to conform to the technology.”

It’s a clear principle: technology should adapt to people, not the other way around.

But with AI, an interesting tension appears. The system is powerful, generates fast results, and often visually good ones… but that doesn’t mean it truly understands people.

“AI today can design the interaction, but without guidance from an expert Interaction Designer… the interface will look wonderful, yet fail to reflect how individual people behave.”

This aligns closely with what we’re experiencing. AI can generate, but without human judgment, conscious iteration, and direction… the result can look correct and still fail where it matters.

And this is where I think that misplaced guilt appears. The system isn’t perfect. But instead of questioning it, we question ourselves.

More responsibility, not less

From the outside, it might seem like everything is easier. And in part, it is. Many tasks that used to take time are now solved quickly. But something changes in parallel: responsibility.

Before, if something failed, it could be a technical limitation. Now, many times the mistake comes from not noticing something that didn’t quite fit.

Generating is easy. Evaluating well is not. And that’s where fundamentals matter more than ever. Understanding what you’re doing, even if you’re not writing everything by hand.

So… vibe coding?

I’m not fully convinced by the term. Not because it’s wrong, but because it falls short. It doesn’t capture the depth of the shift. If anything is clear, it’s that the work hasn’t disappeared. It’s just become less visible.

And maybe what we’re seeing is not simplification, but a fairly natural evolution.

Less execution. More judgment. More attention. More responsibility.

And even if it feels strange at first, everything points to this being the new balance.

References and further reading

1. Don Norman — Guest column: I’m a computer scientist
2. The Secret Life of Claude Code: Skill Is Not What You Think It Is
3. Vibe Coding Aesthetic — Coding Principles

Postscript. The second article shifts the perspective quite a bit. It’s very inspiring and honest: through a narrative format, it shows the real skills behind working with AI. We move from passive users to “augmented creators,” and it clearly hints at where our careers might be heading.

The third article stood out to me because of its lighter, less anxious approach. It doesn’t try to define things rigidly, but rather observes what’s happening and how it feels to work this way. It complements Don Norman’s more critical view with a fresher, more current perspective.

EBook

📕 For a deeper, more practical take on UX, you can read UX for Real — an honest, experience-based guide for designers in their first years, including 50+ UX terms explained in a simple way. Available on **Amazon and [Gumroad](https://kavitauxdesign.gumroad.com/l/ux-for-real)**.