Rethinking Our Relationship with AI

We’ve all had that moment—sitting in front of a prompt box, wondering if we’re “doing it right.” With AI tools at our fingertips, the temptation is to treat them like vending machines: input a request, wait for a satisfying output, move on. But what if we’re missing the bigger opportunity? What if these tools weren’t just utilities, but actual teammates in our collaborative process?

When we proposed a workshop for South by Southwest, we knew we didn’t want to stick to surface-level tips or demos. The organizers challenged us to go deeper—beyond the AI hype and toward something more transformative. And so, we leaned into a question that’s becoming central to our practice: What does it look like to truly collaborate with AI? Not just one-on-one, but as part of a team—many-to-many, in real time, across disciplines.

This was the genesis of our “AI Teammates” workshop. Drawing from tried-and-true facilitation techniques, we reimagined AI as a participant in the room. From the very start, we wanted to shift how people perceive their relationship with these tools. It’s not about asking better questions—it’s about asking better questions together.

The results were electric. People didn’t just learn—they transformed how they thought about AI. They saw themselves not just as users, but as facilitators of AI conversations. They began to glimpse a future where AI isn’t separate from our teams, but embedded within them.

Reimagining AI Through Personas

To help people enter this new headspace, we began with something deceptively simple: a set of AI persona tarot cards. These weren’t just warmups—they were intention-setting tools, designed to spark self-awareness and curiosity. Participants drew cards representing different AI roles: the Challenger, the Historian, the Synthesizer, and the Optimist. Each represented a style of interaction and insight.

From there, we asked them to reflect on the following: Which persona reflects how you currently use AI? The answers served as mirrors, revealing habits and blind spots. Some noticed their go-to AI interactions leaned heavily into optimism, while others hadn’t thought to use the AI as a Challenger or a Historian. The room started to buzz—not just with conversation, but with realization.

This exercise wasn’t just about introspection. It created a shared language for teams to explore how they engaged with AI. Suddenly, AI wasn’t a black box or a mystery. It had personality. It had range. It could wear different hats depending on what the team needed.

And here’s the kicker: People started seeing their own biases and styles more clearly through how they prompted the AI. They also began considering what perspectives were missing. In a team stacked with Optimists, who’s playing the role of Devil’s Advocate? That insight alone sparked new dynamics in the way teams used AI throughout the rest of the workshop.

Prompting as a Team Sport

Once personas were in play, the real fun began. We invited participants to explore a real-world organizational challenge through the lens of their AI teammate. What happens when you tackle a problem with a Challenger AI? How does the response shift when your AI wears the Historian’s hat?

We watched as teams began tweaking their prompts—not just once, but iteratively. “Let me try that from the Synthesizer’s angle,” one participant said. Another team noticed their initial question had been too narrow and asked, “What would an Optimist say if they were trying to pitch this idea to a skeptical executive?”

This prompted a new layer of collaboration—not just between human and AI, but between teammates. People began co-designing prompts, inspired by each other’s strategies and observations. Some even started using different AI platforms (ChatGPT, Gemini, Perplexity) and feeding responses into a shared Miro board, where the ideas could be synthesized and built upon collectively.

This iterative cycle—prompt, reflect, remix—became the heartbeat of the session. It wasn’t about finding the “right” question. It was about evolving the conversation in response to the group’s curiosity. And that, more than any technical breakthrough, is the mindset shift we’re after when we talk about AI teaming.

Bringing AI into the Room—Literally

Most people think of AI as a tool you consult before or after a meeting. But what if you could bring AI into the room in real-time? In the second half of the workshop, we pushed participants to imagine AI not just as a participant, but as a co-facilitator.

Here’s how it worked: We gave them a live meeting scenario—something going off the rails—and asked them to prompt the AI for in-the-moment facilitation help. Not “What’s a good agenda?” but “What should we do for the next 15 minutes to bring this meeting back on track based on our original purpose?”

The responses were surprising, creative, and sometimes hilariously off-base—which made for rich team discussions. But what mattered most was the shift. Teams were engaging with the AI in the moment, treating it not as a scribe or planner, but as a facilitator. They were inviting the AI into real-time problem-solving—just like they would with any other team member.

And because each participant had a different persona in mind, the diversity of responses grew exponentially. One team’s Challenger AI might poke holes in a proposed solution while another’s Synthesizer AI tried to weave together contrasting ideas. And all of this was visible in real-time on the Miro board, where teams could compare notes, build on each other’s work, and generate collective insights.

The Power of Group Sensemaking

As the teams worked, something incredible started to happen: collective intelligence took center stage. The room became a living organism—AI prompts feeding human insight, which then sparked new prompts, which then seeded even richer responses.

This wasn’t just about AI being smart. It was about humans working smarter with AI. People were teaming not just with AI, but with each other—through AI. It was a case study in group sensemaking, powered by diverse perspectives and iterative prompts.

At one point, we noticed teams prompting with an eye toward others’ personas. “I’m usually an Optimist, but let me try this like a Historian.” That cross-pollination of thinking styles is hard enough with human teammates. Seeing it happen with AI added an entirely new dimension.

We even saw people assigning different roles to different AI platforms—using NotebookLM for document summarization, ChatGPT for brainstorming, and image generators for visual exploration. It was like assembling a team of AI specialists, each with a job to do. And the team—the human team—was coordinating it all in real-time.

From SME to Creative Collaborator

In the final phase of the workshop, we introduced a new prompt: What if AI could serve as a Subject Matter Expert (SME)? We gave teams common roles like product manager, designer, or engineer, and asked them to prompt AI to identify what perspectives were missing from their project.

The results were astounding. Participants uncovered blind spots they hadn’t considered. Some even had visceral reactions—one participant who worked in AI said they had goosebumps thinking about how their tools might evolve. AI wasn’t just helping solve problems. It was helping reframe them.

We also played with advanced tools like Miro Sidekicks, which allowed us to synthesize insights from participants’ sticky notes in real-time. We ended with a classic facilitation activity: “I used to think… now I think.” Participants entered their reflections into the board, and Sidekicks turned them into key themes and next steps.

This real-time group reflection—facilitated by both human and AI—offered a powerful closing moment. Teams could see not just how their thinking had changed, but how collective reflection with AI could accelerate learning, deepen insight, and spark new directions.

Using AI to Expand Classic Facilitation Exercises

One surprising outcome of this workshop was discovering how AI could enhance exercises we’ve been using for years. Take the brand takeover activity. Traditionally, we’d assign groups a brand like Nike, Disney, or Chanel, and ask them how that brand might solve their current problem.

Now, using AI, even a team of three can get rich results. Ask AI to roleplay as a Nike strategist and boom—philosophies, playbooks, tone, and style all pop into view. Then ask AI to roleplay Apple, and suddenly you’re switching lenses with ease.

This not only accelerates the activity—it enriches it. Teams can prompt AI to generate visuals, slogans, or mock ads. And even if the outputs are flawed (hello, six-fingered hands), they often spark brilliant ideas. The hallucinations become a feature, not a bug.

Better still, you can have AI personas debate each other. What would Nike and Disney build together? What if Chanel redesigned a Nike product? This “AI team of teams” idea turns solo brainstorming into a rich, multi-perspective dialogue—and invites facilitators to orchestrate that dialogue like a symphony.

Closing: The Future Is Teaming

We ended our workshop—and this reflection—with a simple idea: the more we treat AI like a teammate, the more value we get. Not by anthropomorphizing the tool, but by engaging with it collaboratively, curiously, and creatively.

Whether you’re prompting pre-meeting, mid-discussion, or during synthesis, your mindset matters. Are you just asking for answers? Or are you asking AI to think with you, alongside others? When we shift from one-to-one use to many-to-many collaboration, we tap into AI’s potential as a real force multiplier.

So here’s your call to action: Start small. Give your AI a role. Try a persona prompt. Run a brand takeover with your team and invite AI into the process. Use tools like Miro Sidekicks to synthesize group thinking. Play, reflect, remix. Because the future isn’t just about using AI.

It’s about teaming with it.