A designer named Jenny Wen went on a podcast recently and said the design process is dead. Somebody dropped the YouTube link into the Oracle AI design Slack and within an hour everyone had watched it. Everyone had a take.

Mocking has gone from 70% of a designer's time to maybe 30%. Engineers are shipping prototypes with AI faster than designers can review them, let alone block them. She describes who thrives in this new world: the strong generalist, the deep specialist, the eager new grad who hasn't calcified yet. Ship fast. Iterate in public. Build trust through speed.

She's right. And I know she's right because I was part of what she's burying.

I spent years being the friction in the design process. Not on purpose — or at least not in a way I could see at the time. Slow the work down. Another round of review. One more exploration. If you weren't slow, you weren't thorough. If you weren't thorough, you weren't serious.

I only recognized this about six months ago.

Before the slow years, I was fast. I spent my 20s and 30s jumping from video editing to motion graphics to web development to Flash. Every jump was instinct. Ride the wave, keep your center of gravity broad enough that you don't drown when one thing disappears.

I've been calling myself a creative generalist for 30 years. That label has paid off every time the industry shifted, because the generalist is always already moving when the specialists are still mourning what they lost. Jenny frames the generalist as a new archetype for the AI era. It's the oldest one I know. AI just finally made it legible — because when the making is cheap, the person who can tell you which option is right and why, across disciplines, turns out to be the one everyone needs. We've been here the whole time. The value was in the connective tissue, and nobody notices connective tissue until something tears.

But agency speed had its own problems. When I brought it into healthcare enterprise software, I kept missing things. The problems were just deeper than anything I'd been trained to see. So I learned to slow down. The enterprise process rewarded it. And eventually the slowness became the habit, the habit became the identity, and I stopped asking whether it was protecting anything or just existing.

Then AI showed up and I had to break every one of those habits faster than I've ever broken anything.

The waves I rode before: Web 2.0. Social media. Multi-platform experiences. Every previous shift had a direction — you could see where it was headed and lean into it.

This one doesn't. AI is moving in every direction at once, and there's an anxiety in sprinting when you can't see a horizon line. I'm learning (slowly, which is ironic) that maybe it doesn't need one. Because if you're wrong at this speed, course-correcting is also fast. The old slow made mistakes expensive. Months lost. The new fast made mistakes cheap. Days, maybe hours.

That changes the math on everything. But it doesn't change the feeling. I'm still the person who hates being wrong, even when wrong barely costs anything anymore.

Jenny describes "research preview" like it's a new concept, but companies have been shipping public betas forever. What's new isn't the idea. It's the speed of the cycle. And the speed changes the physics even when it doesn't change the concept.

Jenny's right about almost everything. The process is dead. The generalist survives. Speed wins.

The part I'm still working out — the part I don't think anyone's worked out yet — is what you keep when you shed 30 years of instincts in six months. Which of the slow habits were protecting something real, and which ones were just comfortable. I don't have that answer. I just know not having a clear answer to the question is better than pretending the answer is "go fast."

PS — If this is pulling at something, I want to hear it. napkin@jeremyfuksa.com. Subscribe here.