A senior colleague found a model eighty percent better than anything she'd used all year. Better at what, though? And in your own work, how would you know if the next version quietly lost it? You work through the same chat box daily with almost no view of what the model behind it can do.

A senior colleague found the best model she'd used all year. A day later, an export directive switched it off for everyone. She went back to Opus and kept working. But what happens when the model that vanishes isn't helping a person, and is instead running a process the business can't switch off?

The most capable AI model on the market just disappeared overnight, and the conversation is all about power. But how powerful the model is matters far less than whether you can tell good output from bad. And that gap doesn't close with better models. It widens.

AI is being sold as a uniformly transformative force. The reality is more nuanced. The most impressive gains come from one domain, driven by years of focused engineering. Here's what that means for your AI strategy and the questions every leader should be asking.

A project needs product data from another part of the organization. The data exists, but the definitions don't match. So the project team fixes it: just enough, just for this use case. Each fix makes perfect sense in isolation. From an organizational perspective, it's penny-wise, pound-foolish.

Data management has a predictable failure mode: lack of urgency at strategy level, lack of clarity in execution, lack of mandate with the fixers. AI is about to make that dysfunction impossible to survive. The answer is not more governance. It is fewer people in the loop.

The data professionals who consistently deliver outcomes aren't the ones with the strongest technical skills. They're the ones who understand how value moves through the organization. Most data problems are adaptive challenges that organizations keep treating with technical fixes.

A few years ago, I watched a product recommender start behaving oddly. The model was fine. The pipelines were fine. What had changed was something nobody on the data science team knew about. That moment showed me something I now see everywhere: data visibility is AI's most overlooked bottleneck.

AI is removing the technical barriers that shaped data strategy. When every team can build, the bottleneck shifts from capability to direction. The data leader's most important job isn't enabling more use cases. It's translating business strategy into clear data priorities and saying no to the rest.

In nearly twenty years working with data, I have yet to meet a business stakeholder who woke up wanting a data product. The shift we need is from describing what data is to articulating what data does. "Data-as-a-product" was the right vocabulary for the wrong conversation.

Someone asks: 'What data do we already have?' Within minutes, the room is problem-solving. Everyone wants to deliver. So the team works with what's available. And from that moment, the most expensive decision in the project has already been made.

Most AI thought leadership falls into two camps: hype or prediction. But there's a more relevant frame: the economics of viability. What's currently too expensive to do in your industry that becomes a real option when AI changes the cost equation? And what does that unlock?

Most governance frameworks grew by accumulation. The 'cut in half' question forces you to explain why each piece exists. What you'll find is which parts have a clear rationale and which parts exist purely because they seemed responsible at the time.

We built models to predict grocery baskets. The better they got, the smaller the baskets became. Google, Meta, and now OpenAI have tried to bolt shopping onto their platforms. The disruption never lands. The gap is not compute power. It is a misunderstanding of how people actually shop.

Most organizations can't answer a simple question: what are you actually optimizing for? With agentic AI, leaving it unanswered has consequences that are faster, bigger, and harder to reverse than anyone anticipated.