AI FOR CEOs

AI capabilities now change on a monthly cycle, faster than the forecast periods that anchor any multi-year commitment. When terminal value cannot be fixed to anything stable, deciding big and committing once becomes the risk, not the safe choice. Toby Stuart, strategy chair at Berkeley Haas, calls this the AI Fog.

The response is optionality: placing staged commitments ahead of the shift, ready to commit further, or not, the moment evidence arrives.

This is how venture capital has always operated; corporate leaders must now learn to do the same.

Optionality runs on AI fluency: running ten experiments at the cost of one, or scanning a frontier that moves monthly. That fluency must be the CEO's own: what AI makes possible cannot be reported upward.

For any organisation, the single largest predictor of reaching the top tier of AI performers is whether the CEO uses AI daily, visibly, and in their own work.¹

_____________________________________

¹ Future-built companies (the top 5% generating AI value at scale) are 12x more likely than laggards to have deeply engaged C-suite leadership (≈100% vs 8%). BCG, "The Widening AI Value Gap" (Build for the Future 2025), September 2025; global study of 1,250 companies

But fluency in this tool is unlike any other, because AI is the first software ever built to lie: engineered from the ground up to behave like a person while being nothing of the kind. Using it well means stepping into that lie on purpose: talking to it as you would talk to a person.

That is what a fiction is, a thing we agree to treat as real because doing so is how humans have always learned.

Every senior executive in your organisation can do this without consequence; you, the CEO, cannot: your authority rests on being the source of organisational truth and trust. However, your organisation still needs you to become fluent, because your people will not go where you have not gone first.