Speed Is the Feature
Why AI-powered making feels magical and why design keeps misdiagnosing the problem
Work that used to take weeks appears in days. Sometimes hours. Ideas move from concept to something running before anyone has time to argue about them. The speed feels unreal, almost suspicious. And the first reaction from design is often caution disguised as critique.
“This feels rushed.”
“This skips important steps.”
“This can’t possibly be high quality.”
I understand the instinct. I’ve felt it myself.
But I think we’re misnaming the moment.
The problem isn’t that things are moving too fast. The problem is that we’re still evaluating this new way of making with tools and mental models that were built for something much slower.
AI-assisted building didn’t just make production faster. It changed the economics of making. When the cost of exploring an idea drops dramatically, speed stops being a tradeoff and starts being a capability.
That’s the part that feels like magic.
And magic makes people nervous.
For decades, design used time as a proxy for quality. Time to research. Time to iterate. Time to refine. Slowness signaled care. Speed signaled recklessness. That equation mostly worked when building itself was expensive.
Now the cost curve has shifted.
When ideas are cheap to test, the question isn’t whether something was made quickly. It’s whether anyone exercised judgment along the way. Speed doesn’t remove the need for design. It removes the excuse to avoid decisions.
This is why “vibe coding” gets misunderstood.
People frame it as improvisation or lack of rigor. In practice, what it exposes is whether a team has alignment, taste, and shared standards. When building is slow, misalignment hides behind process. When building is fast, it shows up immediately.
This is where visual design evolves.
Not as polish. Not as surface. But as signal.
When you can generate ten versions instantly, the value is no longer in producing options. It’s in knowing which ones feel right and why. Taste becomes a form of leadership. Visual decisions stop being aesthetic preferences and start acting as constraints the system builds against.
Discernment becomes the work.
Designers are no longer just shaping what things look like. They’re shaping what gets allowed. What gets rejected. What gets refined and what gets shipped as-is. In fast systems, taste is the difference between coherence and noise.
The uncomfortable truth is that many design rituals were doing quiet work we didn’t name. They forced conversation. They slowed teams down just enough to surface disagreement. When AI collapses those steps, we feel the loss and mistake it for a quality problem.
It’s not a quality problem.
It’s a coordination problem.
Design’s role doesn’t disappear in fast environments. It sharpens.
Instead of producing artifacts as proof of thinking, designers are now asked to provide direction that machines can execute against. That means articulating what good looks like before the pixels appear. Naming constraints early. Making taste legible, not subjective.
The cost we should be talking about isn’t speed.
It’s what happens when speed arrives before clarity.
When teams move fast without shared intent, they don’t ship better products. They ship louder ones. More features. More surface area. More decisions made implicitly instead of deliberately.
This is where design needs to evolve.
Not by slowing things down artificially, but by moving upstream. By making judgment explicit. By turning values, principles, and visual standards into inputs, not post-hoc critiques.
If making is cheap now, then discernment is the scarce resource.
That’s not a threat to design.
It’s the evolution of it.
Putting This Into Practice: Designing for Speed Without Losing Taste
If you’re working in an environment where AI-assisted building accelerates everything, the goal isn’t to reintroduce friction. It’s to decide where judgment lives.
Define what “good” is early
Before prompts, before prototypes, write a short statement answering:
What should this feel like when it’s working?
What would make this feel wrong, even if it technically works?
What kind of taste does this product require?
This isn’t a style guide. It’s a north star.
Use fast prototypes as mirrors
When something gets built quickly, don’t judge it as a final artifact.
Ask instead:
What does this reveal about our assumptions?
Where does it drift from our taste?
What did the system optimize for that we didn’t intend?
Speed surfaces misalignment. Let it.
Make taste discussable
Taste doesn’t scale when it lives in one person’s head.
Designers can help by naming what feels off and why. Not with defensiveness, but with language:
“This technically works, but it breaks our sense of calm.”
“This solves the task, but it undermines trust.”
“This looks right, but it feels wrong.”
These are design decisions, not opinions.
Decide where slow is sacred
Not everything deserves deep refinement. Some things absolutely do.
Identify where taste and care matter most:
Trust moments
First impressions
Irreversible actions
Protect time there. Let the rest move fast.
Speed didn’t break design.
It revealed that visual design, taste, and discernment were never optional. They were just harder to see when making was slow.
The future of design isn’t about resisting fast tools.
It’s about leading with judgment when making is no longer the hard part.
Further reading from leaders I admire
- Benhur Senabathi, *Designers as Agent Orchestrators* — UX Collective
- David Mitev, “Quiet Shift Changing UX”
- David Armano, writing on intelligent experiences