The real AI divide isn’t access. It’s who’s still doing the thinking.

People are weaving AI into every corner of their lives. Not just for productivity, but for professional decisions, emotional support, even health advice. That might sound like progress, but something quieter is happening underneath.

When we outsource our thinking, we stop building the capability that comes from actually solving problems. Skill is built through practice and friction. The struggle of working something out is what makes it stick. When we reach for AI before we’ve even tried, we skip that part. We become more dependent and less capable, and because the decline is gradual, we’re the last ones to notice.

That’s the risk. But the same tools are also creating the biggest opportunity I’ve seen in two decades of strategy work. The difference between the two comes down to one thing: whether you bring real thinking to the tool, or use it to avoid thinking altogether.

The future of work has already changed shape.

Over the next few years, we’ll see significant fragmentation. Micro teams, solo operators and small outfits will do things better, faster and cheaper than their bloated predecessors, all of them powered by AI.

This is Jevons’ paradox in motion. When technology makes something cheaper, demand doesn’t fall. It grows. Expertise is getting cheaper to deliver, so more businesses will buy it, just not from the same people. The large institutions built on layers and overhead will struggle to justify their price tags. Consulting, advertising and research are first in line.

I recently watched a $5 million media account move from a global agency to a small media shop nobody had heard of. They won it by thinking differently and delivering more value, without the layers of approval, generic strategy and institutional inertia that come with scale. The client noticed.

Why the small teams are winning.

Big businesses are naturally cautious. Trust takes time. But when they give it, the return is disproportionate.

Rather than paying an agency half a million dollars for a brand campaign, a marketer can now bring in senior strategy and creative talent as a micro team. A strategist and a creative director, for example, partnered with a production company, for around 20% of that budget. The same quality of thinking at a fraction of the overhead.

I ran a two-hour workshop with a startup recently. By the end of the session, they had complete clarity on their product direction. Within hours of leaving, they signed their second ever client. In a traditional agency engagement, with me doing the same work, that clarity would have taken two to six months and tens of thousands of dollars.

Two hours.

But here’s the part that matters. The AI didn’t create that clarity. Twenty years of strategy experience did. The tools turned the session into deliverables they could act on the same day, instead of a report that gets filed away. AI multiplied what I already knew. It couldn’t have replaced it.

That’s the point the opening risk and this opportunity have in common. AI rewards the people who have done the thinking and punishes the people who use it to skip the thinking.

The people most at risk are the ones who feel safest.

There’s a well documented bias called the ostrich effect, our tendency to avoid information that might be threatening. A lot of people are living inside it right now. They know AI is changing things. They’d just rather not think too hard about what it means for them.

AI is an undercurrent, not a wave. It isn’t crashing over us. It’s quietly reshaping everything beneath the surface. The people left behind won’t be the ones who lost a competition. They’ll be the ones who opted out of a shift that was happening whether they engaged or not.

If you’re a data analyst, a strategist, a researcher or a lawyer, the question isn’t whether AI will affect your role. It’s whether you’ll be the person in the room who knows how to use it well, or the person watching someone else do your job faster.

The choice, while it’s still yours to make, is simple. Keep doing the thinking. Then let AI make it go further.