Introducing the AI systems designer

Companies are going to start hiring “AI systems designers” soon.

And the more I think about what that role might entail, the more excited I get about it.

Here’s my thinking on what it might look like 👇

The role will be horizontal

Like design systems, growth design, etc… AI systems designers will span the organization and touch every part of the product.

Right now, AI is mostly contained to chat UI that is layered on top of the core UX. But not for long…

“AI is going to become so deeply woven into the fabric of our products”
— Jordan Singer

You can see a world where more products have AI baked into the interface layer where each affordance seamlessly feeds context to an underlying model. This enables AI to become much more proactive (think: suggestions, simple actions, dynamic dashboards, etc.)

In that world, effective systems design becomes essential.

Because AI features can quickly start to feel like cookie banners if not implemented tastefully 😬

Someone needs to work cross-functionally to ensure every implementation helps users establish mental models for how to interact with AI. Building familiarity will be key.

“As we add new features it should feel the same and not feel like it came from a different team, but rather it came from Figma”
— Jordan Singer

The role will revolve around education

In the same way that design systems designers teach people about what components they have at their disposal, how they work, etc…

AI systems designers will educate product designers about AI’s capabilities (potentially creating totally new types of internal documentation).

All of a sudden, AI is a major variable to consider when imagining potential futures.

Someone has to ensure the team understands:

1. how high the ceiling is for design

2. how AI can help achieve user goals

3. how designing with AI fits into brand guidelines

The role will be strategic

I asked Dan Mall how he measured success for design teams. His answer surprised me:

“It’s that the design system team gets invited to product team kickoffs earlier”
— Dan Mall

I don’t think AI systems designers will struggle to make this kind of strategic impact.

Because product teams everywhere are trying to figure out how to make the most of this new technology. Many companies need to figure out how to leverage AI or risk going extinct.

That means AI systems designers will play a major role in future-proofing the company.

They’re the ones who are responsible for anticipating future trends and helping their team adapt as AI capabilities increase exponentially. It’s not just about understanding how the technology works. The key is identifying how it unlocks design.

The role will require prototyping

Perhaps the primary way this new role will differ from design systems designers is that it will involve a lot of prototyping.

“If I'm thinking about some crazy idea, my process is to hack my way through the Figma code base, tack on some little AI feature and try to get people excited about it.”
— Jordan Singer

Functional prototypes will become way more important when designing with AI.

Because the quality of an AI interaction is directly tied to what the model outputs.

That’s why AI systems design will require some technical chops. You don’t need to be a full-blown engineer, but you do need to be comfortable rolling up your sleeves and hacking on your ideas.

The role will become vital with dynamic interfaces

Soon the product we "ship" might feel more like the base structure or a tailored set of defaults. Someone is going to have to establish guardrails at a system level to ensure AI can spin up dynamic interfaces around it.

I'm not sure what it will entail exactly yet… but empowering AI to effectively wield (and extrapolate) your company's design system will become an important function of this new role.

The future of AI in Figma

In many ways, Jordan Singer’s new role at Figma is a prototype for what an AI systems designer might look like. That’s why I was so excited to interview him live at Config 👀

Our chat follows his journey from the Diagram acquisition all the way to Config 2024. In addition to breaking down his new AI systems role at Figma, we go deep into:

• What didn’t ship at Config and why

• How they navigated the early idea maze

• How the role of designer will be impacted by AI

• What it will be like designing dynamic interfaces

• Why renaming layers was such a big internal debate

• Why the sparkle icon will become the new floppy disk

• The surprising pivot that led to Jordan’s favorite AI feature

• How Jordan prototypes by hacking on top of the Figma code base

Listen on YouTube, Spotify, Apple, or wherever you get your podcasts 👇