AI as Design Material

From Plywood to Prompts: The Evolution of Material Thinking in Design

Design has always evolved hand-in-hand with material innovation — whether shaping wood, steel, fiberglass, or pixels. In 1940, at the Cranbrook Academy of Art, Charles Eames and his friend Eero Saarinen collaborated on MoMA’s Organic Design in Home Furnishings competition. Their submission wasn’t just a new idea in form—it was a new way of making furniture, using innovative manufacturing methods to mold plywood into complex, organic curves.

Plywood wasn’t new. But how they worked with it was. By layering thin veneers, alternating wood grain directions, and applying carefully calibrated heat and pressure, they bent a centuries-old material into shapes never before seen in mass-produced furniture. This process—iterative, tactile, grounded in experimentation—laid the foundation for what Charles and Ray Eames would later refine with the invention of the Kazam! Machine.

For the Eameses, design was problem-solving. It required engineering, craft, and a deep relationship with the material at hand. Drawing the curve of a chair was only the beginning—real design happened in the glue, the mold, the trial, the error.

And materials evolve.

Where once designers shaped wood, metal, or fiberglass, today we engage with an entirely different kind of material: Generative Artificial Intelligence. Unlike traditional materials, AI is intangible — made not of grain or fiber, but of algorithms and probabilistic models. It responds to input with code, images, and text.

Just as designers once shifted from working with raw materials to working with computation, the notion of materiality expanded to include digital experiences, visual fidelity, and the shaping of information. With each shift, design found new ways to understand and manipulate the capabilities and constraints of the medium.

AI has those very same properties — capabilities, constraints, affordances — but it is rarely treated as a material. We tend to see it as “intelligence,” something magical or autonomous. But the true power of design lies in reframing AI — not as intelligence, but as material. As tool.

This evolution mirrors previous moments in design history. In the 1980s, Bill Moggridge and Bill Verplank coined the term Interaction Design to describe how designers were beginning to shape not just objects, but digital behavior. Later, Don Norman helped popularize the term User Experience, broadening the field of design to encompass not just interfaces, but the entire human interaction with a product or system.

By the early 2000s, places like IDEO, the Stanford d.School, and the Interaction Design Institute Ivrea treated computing as a core design medium. Tools like Arduino and Wiring emerged—not just as new hardware platforms, but as vehicles for democratizing computation. Designers began to prototype directly with code and sensors, exploring how software and hardware could shape behavior.

As these tools evolved, so did our understanding of their limits. Constraints — like sensor accuracy, processing speed, and interface fidelity — became the very materials we designed with.

Now, AI continues that evolution. But instead of typing code, we converse with the machine. We prompt. We iterate. We collaborate. Tools like ChatGPT, Midjourney, Rosebud, and SimTheory give rise to a new class of creators—people who may not write scripts or model 3D objects, but who can describe a game, an idea, a story—and see it emerge.

I’ve started seeing this shift firsthand. At GDC 2025, during a Gaming AI panel hosted by EPAM, we discussed the rise of what I’d call "Citizen Game Makers". These are not AAA developers or indie teams, nor are they UGC creators. They’re people using AI to prototype play in real time. With a single sentence, they generate interactive experiences — imperfect, but immediate, and with the potential to be deeply personal and personalized.

Even at home, this shift is tangible. My daughter and I have begun designing simple games together. We talk, we prompt, we test. And in that process, the AI becomes almost like a third member of the design team—not just executing commands, but filling in gaps, making decisions, interpreting nuance.

Prototyping with AI: Designing in Conversation

“The designer imagines the form and behavior of a product, describing how it should look and act [...] Each new technology becomes a tool in the hands of the designer, expanding the possibilities for communication and interaction.” Bill Moggridge, Designing Interactions

Traditional prototyping involves sketching, coding, or building by hand. AI, by contrast, invites us to co-create with a system that makes inferences, introduces variation, and—delightfully or confoundingly—sometimes misunderstands us. This is not just about accelerating workflows. It’s about engaging with a material that pushes back. The resulting artifacts—apps, games, webpages—may not be exactly what we envisioned, but they are explorations of ideas.

At IDEO, there’s a well-known story that illustrates the power of prototyping to align a team. A surgeon had come to discuss a new idea for a handheld surgical tool. For hours, the team talked—imagining, sketching, and debating the form and function. Then one designer stood up, gathered bits and pieces from around the room—a marker, a clip, some tape—and quickly assembled a rough prototype.

Holding it up, they asked: “You mean like this?”

“Exactly!” the surgeon replied.

Prototypes are tools for conversation, agreement, and refinement. Talking about an idea keeps it abstract. Building something — anything — brings it into focus.

In traditional workflows, a prototype shows what you already imagine. But with AI, the prototype becomes a generator of new questions. It suggests patterns, fills in blanks, and introduces unexpected directions — helping you see your idea from a new angle.

Examples:

• A game designer prompts an AI to create a dungeon layout — and receives something surprising, sparking new gameplay mechanics.
• A visual artist describes a character in words, and the AI captures a mood they hadn’t considered.
• A product team rapidly explores different onboarding flows — not to find one perfect answer, but to expand the range of what’s possible.

These aren’t just faster iterations. They’re new kinds of collaboration.

We’re watching the foundations of AI design tooling take shape in real time. Technologies like MCP and advanced models — Claude 3.7, Gemini 2.5, GPT-4o, DeepSeek — are shaping the next creative frontier.

As these systems become more capable and accessible, the real question becomes: Who will design the next generation of tools for this emerging mode of conversational design?

When we treat AI as a design material, prototyping becomes less about refining known ideas — and more about expanding the space of what’s possible. It’s messy, surprising, sometimes frustrating — but that’s what working with any material feels like in its early days.

Clay resists. Wood splinters. AI misinterprets.

But in that material friction, design happens.

The challenge ahead isn’t just to use AI more efficiently — it’s to foster a culture of design experimentation around it. Like any great material, AI won’t reveal its potential through control, but through play, feedback, and iteration.

Now step into the workshop...

If AI is a new design material, the best way to understand it is to start working with it —sketching, poking, bending, breaking. You don’t need to build a product. Just experiment.

Here is another sketching session we just did...

And try out the game here! https://simulationtheory.ai/c659c73b-4cfc-45da-b1b1-6230f16f863f

Some tools I've been using include: Simtheory, ChatGPT, Copilot, Rosebud...

Please add your experiments, prompts and surprises in the comments...