No One Is an AI Expert (And That's Okay)

A Field Guide for CEOs in the AI Era

It's the question echoing through every boardroom in 2025:

"What's our AI strategy?"

If you're a CEO, you've heard this question more times than you can count. It's asked by your board, your investors, your executive team, and sometimes even your customers. The pressure to sound decisive and visionary in response is intense — but the reality behind the question is far messier.

Here's the truth, and I say this with the full weight of having led through two previous technology super-cycles: no one is truly an AI expert. Not yet. Not in the way it matters for building lasting businesses. And that's okay.

We are all operating in a moment of deep uncertainty. The pace of change is breathtaking, the surface area of innovation is expanding by the day, and the temptation to pretend we understand it all — to maintain momentum or authority — is strong. But experience has taught me that pretending is what kills companies.

When I was at Microsoft during the height of the dotcom boom, I saw how dangerous it can be when the pressure to appear "strategic" overrides the need to be grounded. Back then, every executive presentation began and ended with the phrase "internet strategy." It didn't matter if the strategy was meaningful or even logical — what mattered was having one. Many companies threw themselves into poorly conceived web initiatives that added cost, complexity, and little value. When the bubble burst, it wasn't the companies who moved first that survived — it was the ones who moved wisely.

Later, I led product teams at Microsoft and then at Xero during the transition to the cloud. The dynamic was similar. This time, it was cloud and SaaS strategies under scrutiny. Again, some companies reacted with surface-level tactics — migrating for the sake of optics, not outcomes. Others used the cloud as a catalyst to rethink their architecture, their operating rhythms, and their trust models. Those were the companies that went on to lead the way.

And now, here we are again. This time, it's AI.

Everyone wants to know if you have a point of view, a roadmap, or an answer. But the companies that will define this era — just like in the dotcom and cloud waves — will not be the ones who rush to adopt language or tooling. They'll be the ones who treat AI not as a strategy but as a capability. A source of leverage. A quiet, compounding force for operational advantage.

The real opportunity isn't in grand declarations. It's in applying AI to particular business problems where it can remove friction, reduce decision latency, or unlock new sources of value. It's not about saying "we're doing AI" — it's about doing things better, faster, and more responsibly because AI makes it possible.

This is why I continue to focus on applied AI — not as a concept but as a discipline. In a recent post, I argued that the most innovative CEOs today are not racing to build flashy prototypes or announce copilots; they're quietly deploying AI inside internal workflows, customer support backends, and decision-support systems. They're shipping experiments, learning fast, and keeping their attention fixed on ROI — not headlines.

In a follow-up piece, I shared why this moment mirrors the cloud transition in more ways than most people realize. Cloud wasn't just about moving workloads — it was about trust. It was about governance. It was about velocity. AI is all of that, only faster and messier.

In a third post, I opened up about how I've used AI personally — to build knowledge hubs, manage my time, and reduce the cognitive load that piles up when you're leading at scale. These aren't moonshots. They're daily improvements. And cumulatively, they matter more than any keynote-stage demo.

So, where should CEOs begin? Not with a five-year AI vision. Not with a mandate to "integrate LLMs" across the company. Start instead by asking where your organization is slow, manual, or bottlenecked by inconsistent judgment. Look for the workflows where a little automation could go a long way. Start there. Don't try to transform everything. Pick one process. Make it better. Learn something. Then, scale what works.

If the cloud era taught us anything, it's that credibility is earned through delivery, not declarations. Delivery in the AI era will depend not on who can talk the most about generative models — but on who can apply them most meaningfully, quietly, and trustworthily.

You don't need to pretend you're an AI visionary. You must lead with clarity, humility, and intention. And if that means saying, "We don't know yet, but we're learning fast," then you're probably on the right path.

At Bodhi Venture Labs, I work with CEOs who are navigating this exact inflection point. They're being asked for certainty in a moment that offers very little of it. What they need isn't a strategy that sounds good — they need a plan that works, starting with small wins, rapid feedback, and a clear connection to value.

You don't need to be first. You need to be real.

None of us are AI experts. But we've been here before.

And we know how to lead when the path is still being built.

Let's get to work.

‍