Most companies miss the AI turn
There’s a growing tendency to talk about AI as if it were a kind of autonomous force, something that will inevitably sweep through companies and transform them from the inside out. And while it’s perfectly reasonable to think that AI can disrupt entire markets or even reshape whole industries, it’s a mistake to assume that the same logic applies at the level of individual organisations. Markets can change without companies changing. Industries can be disrupted while many of the incumbents remain stubbornly, and sometimes fatally, the same.
That distinction only really clicked for me after a very concrete experience. A few years ago, I worked with a large company and tried to help them automate one of their core processes. Technically, the case was obvious. The process was slow, manual, and riddled with inefficiencies. We spent months trying to move things forward, but over time it became clear that the obstacle wasn’t the technology. It was the organisation itself. They didn’t really understand what the technology enabled, nor were they prepared to rethink how the work should be done. Eventually, I gave up. Two years later, they called me back. AI had advanced, tools were cheaper and easier to use, and the general discourse had shifted. And yet, when I returned, the same work was still being done largely by hand. The market had moved on. The company hadn’t.
That experience made something very clear to me: technology can transform an environment without transforming the actors inside it. AI can redraw competitive landscapes, create new winners, and eliminate old advantages, but it does not, by itself, rewire how a specific company thinks or operates. If the internal culture and ideology aren’t ready, the technology just gets absorbed into existing habits. You don’t get transformation; you get continuity with better branding.
One major reason for this is that senior leadership is often too far removed from the actual work. Managers at the top may know, in an abstract way, that there are bottlenecks, but abstraction smooths out friction. The real inefficiencies live in the details: the handovers, the exceptions, the constant rework, the office politics, the informal fixes people rely on just to keep things running. If you haven’t spent time close to that reality, you don’t form a reliable mental model of how the process actually behaves.
From that distance, decision-making becomes distorted. Leaders issue broad directives about efficiency or innovation, and middle management responds with proposals that are safe, incremental, and politically viable. Middle managers are typically very conservative in their approach to change. The proposals get approved not because they solve the real problem, but because the people approving them don’t have enough grounded understanding to tell the difference. As a result, things change on the surface, while the core constraints remain untouched.
This is where the role of the CEO becomes especially important. At its core, the job requires holding two demanding models in mind at the same time. One is outward-facing: an understanding of the market, customer preferences, competitive dynamics, and how those forces are likely to evolve. That model feeds strategy. The other is inward-facing: a clear picture of how the company actually works, how its processes fit together, where they break down, and how flexible they really are.
If either of these models is weak, the consequences are predictable. A CEO who understands the market but not the organisation can articulate a compelling strategy that never quite materialises. A CEO who understands operations but not the market can run the company very efficiently in the wrong direction. Strategy, in practice, is the act of translating external pressures into internal change. That translation only works if both sides of the equation are well understood.
What AI does, then, is not magically fix this problem. It operates at a different level. It raises the stakes by changing what is possible in the market, not by ensuring that any given company will respond appropriately. Some firms adapt, others hesitate, and many simply carry on until competitive pressure catches up with them.
AI changes the external world, but for a company to succesfully adapt, that change needs to:
be incorporated into the CEO’s mental model of the market;
be translated into the CEO’s mental model of the company;
be translated into clear, detailed top-down policy;
be translated into coherent action at all levels of the org chart.
This information flow presupposes a few critical things, namely that the CEO’s mental model of the market correctly reflects the market, that the CEO’s mental model of the company is actually correct, that the CEO knows how to design and enforce top-down policy and coherent implementation plans and that the politics of running the business align correctly to make things work at all levels. Pictured this way, adjusting a business to an AI-led tech landscape sounds like rocket science. And, honestly, it probably is.