How I Lost My Love For Design And Found It Again

A couple years ago, I realized that somewhere along the way, I stopped designing. After Flexport’s hypergrowth, I moved into management, leading a team of 12. Then I co-founded a startup, joined Y Combinator, raised capital. I was pursuing the dream. But that dream left little time for what I loved most. My days were filled with managing, hiring, fundraising, operations, attorneys, taxes, etc.

At the same time, the tools moved forward without me.

Figma didn’t just improve, it transformed. Auto layout, nested components, and complex variants turned static mocks into living systems. What once required a dozen Sketch files and brittle handoff docs now lived in a single, scalable canvas.

Design became code-adjacent. Then code-aware. Then code-generative.

Just as we learned to work at that new level of fidelity, AI arrived. Not as a distant promise or research paper, but as something real. Large language models put AI in everyone’s hands. The shift was no longer hypothetical. It was happening in production.

AI wasn’t just a faster assistant. It was a new kind of collaborator. It didn’t wait for input. It guessed. It responded. It shaped intent.

Interfaces stopped reacting to clicks and started responding to uncertainty. To goals. To ambiguity. The floor under design started to shift. Most people didn’t notice. Some still don’t.

I watched it happen—first with skepticism, then with awe. Not just because of what was possible, but because of what was at stake.

So I started over.

I took a year to rebuild. I learned Figma again. I taught others through a YouTube series. I built an enterprise-grade design system. I studied AI—not just how it worked, but what it meant. And I realized something important:

We’ve always shaped behavior. That’s what design does. It guides action, influences perception, and helps people make decisions. But now the system behaves too. It interprets. It adapts. It changes with each input. It no longer sits quietly between the user and the output. It acts.

Designers no longer control the experience, they guide it. Today, design means shaping the behavior of intelligent systems. It means making that behavior legible, trustworthy, and valuable. The interface has become dynamic. It no longer follows a linear path. It generates possibilities.

Design Is Changing

Designers used to build precise machines. You could engineer every outcome. You knew exactly what happened when someone clicked.

Today, the work feels different. It’s less like engineering and more like ecology. You plant conditions. You manage inputs. You shape what grows. You don’t handcraft every flow. You shape probabilities. You prompt. You interpret. You adjust the scaffolding and watch what emerges.

If the model drives the engine, the interface steers the experience. If that steering feels wrong, people lose trust. If it disappears, people lose control.

The job now is to help humans work with a system they don’t fully understand. That requires new patterns. New skills. And a new mental model for what design even is.

The Tools Shape the Thinking

Every generation builds inside a different dominant medium. Writing taught people to think linearly. Television trained people to process emotion over argument. The internet broke everything into fragments—links, feeds, tabs, notifications.

Each medium changed how people communicated and how they understood the world. Not just because of the content it carried, but because of the structure it imposed.

Now, the dominant medium is intelligent systems. We don’t navigate software anymore—we describe intent. We don’t click through flows—we interact with models that adapt and respond.

This behavior changes how we think. People now expect outcomes before they understand how those outcomes happen. They expect results to appear through collaboration, not command. That changes how people learn, decide, and trust.

A New Class of Apps

We’re starting to see apps emerge from this new thinking. They don’t use rigid flows. They don’t walk users through ten screens. They ask, “What are you trying to do?” and then try to do it. Notion AI. Perplexity. Replit. ChatGPT. These tools compress entire workflows into a single input.

Paste a spreadsheet, get a chart. Write a sentence, get a campaign. The surface feels thin, but the result is dense. The system does more. The user does less. It feels like magic.

This change is already reshaping expectations, especially in consumer software. Creators build videos, songs, and copy from a few lines of input. The tool disappears. The work accelerates.

But enterprise software hasn’t caught up.

Business tools still run like it’s 2012. Ten-step forms. Deep menus. Static tables. Software still forces users to think like machines. Most AI features feel bolted on, rather than integrated or thoughtful.

That won’t last.

Employees already expect more. They won’t keep clicking through tabs when AI can do the work. They won’t wait on IT tickets when a prompt could handle it instantly. This pressure is both cultural and technical.

The shift is coming fast.

Society Is Already Changing

This transformation reshapes more than just work or tools. It’s changing society. Not in a sudden collapse, but in quiet, steady shifts.

People no longer have to navigate fixed systems. They can negotiate with responsive ones.

This shift is changing how people work, learn, create, and trust. The world will feel normal, until we look back.

In ten years, the past will look unrecognizable. We’re compressing decision-making, coordination, and execution into intelligent systems. They scale instantly. They respond instantly. They carry hidden logic. This change will have consequences—some we can predict. Most we can’t.

But one thing is certain: a sorting is coming.

Between those who adapt and those who don’t. Between companies that redesign from the ground up and those that cling to the past. Between people who embrace this new medium and those who get left behind.

And the shift is already underway.

Are you prepared?