Fork yourself before the market does it for you.

Shell's planners never forecast the embargo. They built scenarios that forced management to live, on paper, in a world of constrained supply and high prices. So when the shock came, Shell recognised it and moved while rivals were still debating whether and how to respond. It came through stronger than competitors who were better at forecasting and slower to react.

That is the lesson for AI.

The instinct now is to predict. What will AI do to my industry by 2030? Wrong question. You cannot forecast a technology that keeps rewriting its own pace. You can only rehearse the scenarios.

And the rehearsal is uncomfortable, because it means asking hard questions about your own business before the market forces them. Does our product remain valuable? What happens to pricing? Do customers still need our category at all? Which part of the value chain still captures the profit? What capabilities would take us too long to build once the scenario becomes obvious?

Sit with those honestly and you find the same thing every time. AI has quietly put every market leader in the same spot. It has set a fork against them.

You make the current model more efficient at the exact moment a challenger uses the same tools to make that model unnecessary. The challenger's move is to find the one thing the leader cannot answer without breaking what made it strong, and AI is now making that move against incumbents everywhere, at once.

So the sharpest move a leader has is to turn the challenger's question on itself, first. If we started today, with AI, how would we make our own business unnecessary? Then build that, before someone else does.

Because disruption rarely kills a company through a wrong forecast. It kills through slow metabolism, the lag between seeing the signal and changing the model. Shell's real edge in 1973 was not foresight. It was reaction time.

You cannot predict AI. You can shorten the gap between the first weak signal and the moment you move.