AI Scenario Workshop

Stop guessing. Start scenario planning.

Your team is anxious about AI. Some see it as opportunity, others as threat. Meanwhile, you're supposed to make strategic decisions without knowing which future will actually happen.

What if you could prepare for multiple futures at once?

Let's Talk →

The Shell Story

In 1981, Shell was the weakest of the Seven Sisters oil companies. By 1990, they were the strongest.

The difference? Scenario planning.

While competitors made single forecasts and got blindsided by oil price collapse, Shell had prepared for multiple futures. When crisis hit, they had a playbook ready.

You face the same choice with AI. Make one bet and hope you're right — or prepare for multiple futures and be ready for what actually happens.

What Happens in a Workshop

2.5 hours. Your team. Four possible futures.

1. Horizon scanning (20 min)

Everyone shares what they see happening in the world of AI — trends, fears, opportunities, weird stuff. We get the full landscape on the board.

2. Tension mapping (20 min)

We identify the real contradictions — the "this could go either way" forces. Your team votes on which tensions matter most. These become the axes of a 2×2 matrix.

3. Build four futures (30 min)

Each quadrant becomes a scenario. Not predictions — possibilities. We give them names, make them vivid, explore what each world would mean for your team.

4. Signals and action (20 min)

What real-world signals would tell you which scenario you're drifting toward? What can you do now that works across multiple futures? You leave with a watchlist and concrete next steps.

The outcome: Shared mental models, reduced anxiety about the unknown, and concrete next steps that work across multiple futures.

Example: Cowboy Engineering Team

In a recent workshop with Cowboy's engineering team, the group explored how AI agent capabilities and society's relationship with truth might evolve. They chose "Fake vs. fact-check world" and "All jobs automated vs. augmented jobs" as their two axes.

The four scenarios that emerged gave the team different lenses for thinking about their product roadmap and priorities — from a dystopian AI-dominated future to one where humans and AI work together with strong verification systems.

Your scenarios will be specific to your uncertainties and context.

I wrote an honest reflection on what went well and what I'd do differently: What I Got Wrong Running My First AI Workshop.

What Participants Say

"It's really the kind of activity that, on top of bringing us closer together, allows us to pause and reflect on crucial aspects of our work and the future."

— Workshop participant, Cowboy

"I'm not a fan of unstructured creative sessions, but I held on and actually enjoyed it."

— Workshop participant, Cowboy

"The most interesting part was when we reflected on the impact of AI on the world, and then translated it into future scenarios for Cowboy. It was very concrete."

— Workshop participant, Cowboy

"I liked trying to see the glass half full by taking fears and flipping them. It's always interesting to take time to pause and reflect on these kinds of things."

— Workshop participant, Cowboy

"I'd recommend this to the whole company — every department can benefit from it."

— Workshop participant, Cowboy

What You Get

  • 2.5-hour facilitated session — remote or Brussels in-person
  • Four documented scenarios specific to your team's context
  • PDF summary of your scenario framework
  • Follow-up check-in to review how scenarios are tracking
  • No pre-work required — just bring your uncertainties

Who This Is For

  • Tech teams (4–15 people) navigating AI strategy
  • Cross-functional groups who need shared language about the future
  • Leaders who are tired of consultants selling certainty they can't deliver

This workshop works because it doesn't require you to predict the future. It requires you to prepare for it.

About the Facilitator

I've spent 500+ hours teaching developers to think in systems at Le Wagon Brussels. Now I facilitate teams to think in futures.

I'm also a software engineer — which means I understand both the technical reality of AI and the human dynamics of teams trying to figure out what to do about it. Scenario planning is complex, but the process shouldn't feel like a PhD seminar.

Format: Remote or Brussels-based

Next step: Email me to discuss your team's context.

george@gkosmo.eu →