The power of AI lies in who uses It

Every June, my brother asks me the same question: ​“Are you joining the Sporza Tourmanager this year?”

For those unfamiliar, Tourmanager is an online game where you build a fantasy cycling team for the Tour de France. With a limited budget, you pick riders who score points based on their real-life performance: stage wins, top 10 finishes, jerseys, and final standings. It’s a game of strategy where true cycling fans get to shine.

Here’s the thing: I’m not a cycling expert. My brother is. Year after year, he crushes me and the rest of our mini-league with ease. Still, I always say ​“yes” when he invites me to join. Deep down, I keep hoping to beat him one day.

This year, I decided to take a different approach. No guessing, no gut feeling. This time, I had a secret weapon: AI.

I use ChatGPT daily at work to test ideas, speed up analyses, and refine strategies. So why not use that same power to win a game? Armed with my digital assistant and a healthy dose of confidence — maybe even a touch of arrogance — I stepped into this year’s virtual Tour.

This was going to be the year where AI helped me win.
Or so I thought.

The AI Strategy: Hoping for a Shortcut to Success

With ChatGPT as my digital sidekick, I kicked off the challenge with full confidence. At work, I’ve seen how AI can speed up processes and sharpen insights — so why wouldn’t it work for a game like the Sporza Tourmanager?

I went all in:

• Data checks: I asked AI to analyze rider form, results from previous Grand Tours, and their roles within their teams.
• Course matching: We reviewed which riders fit each stage and where points could be gained.
• Team composition: Sprinters, climbers, GC contenders… I aimed for the perfect balance.

After several iterations, I had what felt like a winning team. ChatGPT provided suggestions, calculated potential point totals, and helped me make tough decisions. For a moment, I felt like a data-driven team director.

And I didn’t stop there.
During the Tour, I kept feeding new information into AI: who should I start, who should stay on the bench, and which transfers would bring the biggest impact? The answers were logical, structured, and gave me the feeling that everything was under control.

I truly believed: there’s no way this can go wrong.

Expectations were sky-high. This would be my year — the moment where technology would outsmart my brother’s gut instinct. But as I would soon learn, AI only works when you have enough domain knowledge to ask the right questions. That realization hit harder than I expected, and it taught me a valuable lesson.

The Wake-Up Call: Technology Without Expertise Falls Short

As the Tour went on, my confidence stayed high. Every day I refined my decisions using AI’s advice. I trusted the data, the patterns, the simulations. Victory felt within reach.

Then came the final results.
In our mini-league, I finished… third.
My brother? Once again, he took the win without breaking a sweat.
And in the overall ranking? I landed in position 79,848.

It felt almost ironic. I had poured hours into analysis, simulations, and AI-driven decisions — only to be beaten by someone who relied purely on experience and instinct.

That’s when it clicked:
AI can support you, but it doesn’t replace expertise.

The reason is simple. AI can identify patterns and make predictions, but without a deep understanding of the context, those insights stay on the surface. In my case, I lacked the cycling knowledge to interpret the outputs properly. My brother had that intuition. He saw things neither I nor the AI could pick up on.

This isn’t just about a game.
The same principle applies in business. Technology, no matter how advanced, only creates value when the people behind it know how to use it. Without domain knowledge and a critical eye, AI quickly hits its limits.

This realization would shape how I think about AI in every context — from a fantasy cycling league to the real-world challenges we face in our projects.

Losing in Tourmanager might sound trivial, but the insight I gained from it is far from small: AI is powerful, but only in the hands of those who know what they’re doing.

I see this every day in my role as an Agile Coach. AI tools help me analyze faster, spot patterns, and automate repetitive tasks. They make me more productive — but they don’t do the thinking for me. They don’t know which questions matter most, which nuances to consider, or which choices truly move the needle. That direction has to come from me.

The same goes for organizations.
Many companies rush to ​“use AI” because it’s trendy or because they fear falling behind. But without a clear purpose and people who understand the context, AI rarely delivers on its promise. AI can’t design a strategy, set priorities, or read between the lines. Those are human strengths.

From my own experience — both in that cycling game and in countless projects — I’ve learned three key things about using AI effectively:

• Start with the goal, not the tool. Know what problem you’re solving before you bring in technology.
• Invest in knowledge. Your team needs to understand both the potential and the limitations of AI.
• Use AI as a complement, not a replacement. It’s an amplifier for expertise, not a substitute.

Companies that grasp this balance create real value with AI. Those that don’t end up with shiny tools and disappointing outcomes. I learned that the hard way through a game — but the lesson applies everywhere.

How We Combine AI and Expertise at Pàu

At Pàu, we’ve never believed in shortcuts. We believe in people. From the very beginning, we’ve invested heavily in developing our consultants — not as a side effort, but as the core of how we work.

Our chapters — dedicated communities around specific areas of expertise — make sure knowledge doesn’t just sit still. They allow our people to deepen their skills, share experiences, and build a strong foundation together. That expertise is what makes any technology we use truly effective.

AI fits into this philosophy perfectly. We don’t see it as the solution, but as an extra layer that enhances what we already do well. To make that happen, we recently worked with our AI team to define how we should approach AI as an organization. The outcome was clear:

• AI is a tool, not a replacement.
• People remain at the center of everything we do.
• Critical thinking and expertise are non-negotiable.

That’s why we’re rolling out a company-wide AI fundamentals program. Every consultant will learn what AI can (and cannot) do, where the pitfalls lie, and how to use it safely and effectively. After that, we’ll move into more specialized training, tailored to each area of expertise and supported by our Chapter Leads.

This approach ensures that AI doesn’t become an isolated experiment or a gimmick. Instead, it becomes a smart extension of our craftsmanship — helping us create real impact for our clients. Because the magic happens not when AI takes over, but when people use it with purpose and understanding.

The buzz around AI is everywhere. New tools appear every week, each claiming to revolutionize how we work. But here’s the truth: technology alone doesn’t create success. It’s the people behind it who decide the direction, set the priorities, and make the choices that matter.

The real strength lies in the combination.
AI can process data at a scale and speed that humans can’t match. But it lacks the creativity, intuition, and contextual understanding that humans bring to the table. When you merge the two — human expertise with AI capabilities — you get outcomes that neither could achieve alone.

This is why we keep investing in our people. Our mission, Unleash creativity and design to develop a better self, a better society, is rooted in human potential. AI is there to support and accelerate that mission, not replace it.

For organizations, this means:

• Use AI as an accelerator, not a crutch.
• Continue to strengthen human expertise.
• Leverage both together for lasting impact.

The future we believe in is not AI versus human. It’s human with AI—a partnership where each makes the other stronger.