Rethinking Figma in the Age of AI

Every year, Config offers a glimpse into how Figma envisions the future of design. This year, however, I found myself paying less attention to individual feature announcements and much more to a broader question:

What should Figma become in the age of AI?

Over the past two years, AI has fundamentally changed how software is built. Vibe Coding has quickly become part of many teams' daily workflow, naturally raising expectations for design. This year's Config felt less like another product launch and more like an opportunity for Figma to define what AI-native design could become.

Beyond AI Features

The keynote introduced many impressive capabilities—Motion, Shader, Code Layers, and a long list of quality-of-life improvements. As a designer, I genuinely appreciated the craftsmanship behind many of these announcements.

What I was hoping to see, however, wasn't simply more AI features.

I was hoping to see the beginning of what I would call Vibe Design.

Not AI that generates another mockup, but AI that understands an existing product, learns a team's design language, and contributes like another experienced designer on the team. Instead of repeatedly prompting isolated screens, the workflow itself becomes continuous, contextual, and collaborative.

To me, that's the next frontier.

Figma Already Owns the Most Valuable Asset

Much of the discussion around AI in design focuses on Design Systems. While they're certainly important, I think they're only one small part of Figma's real advantage.

The most valuable asset already exists inside every mature Figma workspace.

• Years of production screens.
• Reusable page layouts.
• Interaction patterns.
• Design decisions.
• Product evolution.

For enterprise products especially, very little is designed entirely from scratch. Most new work builds on years of accumulated knowledge. Thousands of files collectively document how a product has evolved, why certain decisions were made, and which patterns have stood the test of time.

If AI can truly understand that context, it wouldn't just generate interfaces. It could design in a way that feels native to the product itself.

My Expectation Was Surprisingly Simple

The workflow I imagined isn't particularly complicated.

I open an existing product and describe the next feature I want to build.

The AI understands the current interface, recognizes established patterns, generates production-ready designs using existing components, and keeps everything synchronized with code while I continue refining the design visually.

• No screenshots.
• No exporting.
• No repeatedly regenerating the same interface.

Just one seamless conversation between the designer, the product, and the AI.

That, to me, is what Vibe Design should feel like.

Why I Left Wanting More

Code Layers is undoubtedly a step in the right direction, and I believe the team is moving toward a much stronger integration between design and engineering.

Still, I left Config wanting more.

Part of that is because products like Paper and Pencil have already changed my expectations. They're still early, but they've demonstrated a vision centered around AI-native workflows rather than AI-assisted features.

I wasn't expecting Figma to imitate those products.

I was hoping Figma would redefine the category.

Instead, this year's keynote felt more like an evolution of today's workflow than a reinvention of tomorrow's.

Optimizing the Wrong Part of the Workflow

One thought stayed with me throughout the keynote.

Many of the announced features help experienced designers create even better work. Motion, Shader, and advanced creative tools undoubtedly raise the ceiling of what's possible inside Figma.

But AI also changes where the biggest opportunity lies.

Helping designers go from 95 to 99 is valuable, but only if teams still choose Figma to get from 0 to 80 in the first place.

The future won't be decided solely by who has the best design features. It will be shaped by who creates the fastest, most seamless, and most context-aware workflow for building products.

Looking Beyond Config

Despite my critiques, I remain optimistic.

Few companies are better positioned than Figma to define AI-native product creation. No other platform possesses such a rich archive of design knowledge accumulated across millions of products and countless iterations. If that knowledge can become AI context rather than static files, Figma has an advantage that's incredibly difficult to replicate.

Beyond the product announcements, my favorite part of Config remains exactly the same every year: the people.

It's one of the few occasions when designers from different companies, cities, and countries gather in the same place. Old friends catch up, new friendships begin, and conversations naturally drift from pixels and prototypes to AI, startups, careers, and where our industry is heading next.

Designers have always been a unique community. We care deeply about details most people never notice—not because of perfectionism for its own sake, but because we believe thoughtful details accumulate into meaningful experiences.

AI will undoubtedly change how we design. It may even redefine what it means to be a designer.

But I hope there will always be a place where people who care about craft, curiosity, and building products with intention can come together.

For me, Config has always been one of those places.

And I sincerely hope it continues to be for many years to come.