How do you use AI to prototype a surgical device? Takashi Wickes’ answer might surprise you.

The interaction designer joined us on the Creative Confidence Podcast to talk about the new IDEO U course he teaches, Prototyping with AI, and share stories of how he’s built AI tools into his creative process over the years.

He recalled a project with an IDEO team where he spent hours talking with ChatGPT to learn how medical procedures worked before interviewing surgeons. AI helped him outline interview questions that drew out the most insightful details, opening up new understanding of the surgeons’ needs and shaping the design of the surgical device. Instead of offloading the design of the device to AI, Takashi and his team used it as a tool to make space for deeper, more productive human conversations.

In this conversation with podcast host Mina Seetharaman, Takashi shares more stories, tips, and frameworks for using AI to bring ideas to life and make the creative process even more human-centered. Together, they explore how to embrace friction, cultivate play, and build the mindsets that keep humans in the driver’s seat.

Listen on Apple, Spotify, or wherever you get your podcasts.

Prototyping with AI: A New Kind of Craft

What is AI prototyping? It’s not just about building AI-powered apps or services. It’s the craft of using AI tools to bring your ideas to life so you can test and iterate on them. You can use AI to prototype almost anything, from physical products to services and experiences.

“We want to bring AI into our process without letting it take over our process,” Takashi explained. “Creativity should always be at the core.”

Instead of focusing on tools and prompts, Takashi emphasizes mindsets for using AI: curiosity, adaptability, and intentionality. Tools will continue to evolve, Takashi said, but a creative mindset will always be relevant.

Be thoughtful about what you delegate to AI and what you keep in your own creative wheelhouse. An unconsidered use of AI would be prompting it to deliver the perfect solution to a brief. Takashi takes the opposite approach—he uses that output as a benchmark. If AI produces the same answer he came to on his own, it’s a sign he hasn’t pushed his thinking far enough.

The better way to bring AI into the process, he says, is to treat it as a “curiosity machine.” It can help you explore more directions than ever before and generate new questions to guide your creativity.

The Power of Intentional Friction

Most conversations about AI center on speed and efficiency. But Takashi makes a bold case for adding intentional friction into the design process.

“Friction allows us to think. It helps us pace ourselves and engage more deeply,” he said. “Without it, we lose the space to absorb and make meaning.”

He shared a story from his time at IDEO Play Lab about a simple online game called Wait, Wait, Wait. The game invites players to move a circle back and forth across the screen. The twist is that each round, the movement slows down. That deliberate friction surprisingly drove higher engagement with the game.

“Friction gives us pacing and a sense of reward,” he said of its application in gaming. “If everything were effortless, we’d lose the value that comes from working through challenge.”

In design, intentional friction works the same way. It helps reclaim time for reflection, inspiration, and human connection. Adding friction might look like scheduling team check-ins to discuss findings or creating intentionally scrappy, low-fidelity prototypes that invite feedback instead of presenting AI-polished renderings.

The Director–Curator–Craftsman Framework

One of the core ideas in IDEO U’s Prototyping with AI course is the Director–Curator–Craftsman Framework, a model that helps people balance human judgment with machine capability when exploring new ideas.

Takashi moves fluidly through these three mindsets when prompting, evaluating outputs, and selecting his preferred AI tools:

• Director – Sets the creative vision and gives AI direction. Like a film director, you decide what story to tell and how to prompt AI with intention.

• Curator – Selects, edits, and makes sense of abundance. The curator’s job is to decide what’s worth showing and what meaning to draw from it.

• Craftsman – Experiments with tools, hones taste, and builds fluency. The craftsman decides which AI tools they’ll use and why.

AI’s ability to produce unlimited ideas can be tempting, but quality matters more than quantity. Takashi advises teams to resist showing every AI-generated option. When you present too many ideas, people often make quick, surface-level choices. Instead, bring fewer, sharper, and more intentional concepts that invite deeper discussion.

“We are curators of a museum that holds every image in the world. If we display everything, it’s meaningless,” Takashi said of the role of humans in evaluating AI outputs. “The human craft is the act of curation—selecting the pieces that matter and juxtaposing them to create meaning.”

Together, these three mindsets form a continuous loop of direction, exploration, and refinement, helping teams stay in control while leaving space for surprise and discovery.

AI as a Playground: Bringing Play Back Into the Creative Process

When AI first entered the workplace, many people felt pressure to master it quickly. Takashi encourages a different approach: one grounded in play.

“A play mindset gives us permission to experiment. You can’t really break AI, so why not try,” he said.

In the course, Takashi encourages learners to treat AI like a playground, entering what he calls the “magic circle” of experimentation for ten minutes at a time, then stepping back to reflect on what they discovered. When you see AI as a space for play, not performance, you start to notice patterns, emotions, and opportunities you’d otherwise miss.

This approach keeps the creative process joyful and human. It allows people to learn and experiment without fear of failure. And it turns AI from a source of pressure into one of confidence and curiosity.

What AI Misses and Why Human Insight Still Matters

Despite its power, AI has its limits. Takashi cautions against relying on it to simulate human behavior or replace user research. While AI can quickly pull from vast amounts of existing data, it only reflects what people have already said—not what they think, feel, or do.

He points to the Say–Think–Feel–Do Framework, a cornerstone of human-centered design, as a reminder that true insight comes from real people and lived experiences.

• Say: What people express out loud
  

• Think: What they believe or assume
  

• Feel: The emotions and tensions behind their words
  

• Do: The actions that reveal their real needs