This is how it starts. Someone puts a word on the screen that costs more than it explains. Today the word is AI. Last year it was digital. Before that, a long time before that, it was computers, the internet, mobile.
The font is big and bold. Charts and bulleted lists with flawless corporate design flow seamlessly across the screen. Lots of approving nods and questions follow. The meeting ends with a plan. Subcommittees, workstreams, timelines and accountability.
The Head of IT has been assigned the role of keeper. New budget. New headcount. AI is now an animal in a zoo. The animal is called AI. The zoo is called strategy. The team can look but the beast can only be touched with a long pole marked security.
Milestone delivered. Your business is the proud owner of an AI strategy. Ambitious, forward looking and useless.
Businesses have been making this mistake for decades. When the iPhone arrived in the corporate world, it spawned an industry called “Bring Your Own Device” or BYOD built to contain the new tech within existing policy boundaries. New companies like Good and MobileIron emerged. Gone and forgotten now.
Too many businesses approach AI with a substitution mindset. What can this do? How would that automate what we do today? That leads to limited and tightly constrained experiments. Can we save a bit of money while avoiding any risk?
Companies that frame their AI problem as a technology problem are answering the wrong question. Your AI strategy is a route map to solving the wrong problem for the next three years.
What is the alternative? There are no certainties. Two things that definitely won’t work. Pursuing a defined solution - avoid anyone offering certainty. And searching for some Holy Grail of top performers - the companies that look smartest now may turn out to be pretty dumb next year or even next week.
Start with the right question: how is AI changing the ground we operate on? Form a view. Try something. Watch what happens. Build on what works. Kill the rest and ask a different question. Evolve.
Caging AI looks like a neat way to allocate scarce resources while everyone else gets on with running the business. It isn’t. You are just pushing the change away.
In order to take advantage of AI, keep asking yourself: What should we stop doing?
“What should we stop doing?” is the right question, Kenny.
One angle I’ve been thinking about: it’s not just strategy vs experimentation, but also how AI fits within decision-making. Most processes are really a sequence of decisions - what data to use, what thresholds to apply, when to escalate.
In many cases, AI gets embedded into these flows while the underlying decision logic remains unchanged.
That’s where things start to fragment - experimentation moves forward, implementation follows, but you end up doing the same things faster.
The harder shift is not adding new capabilities but changing how decisions are made.