AI Isn't The Future - It's the Infrastructure

There’s a lot of noise right now about AI – which tool to use, which model is best, what getting started even means. .

Here’s the truth: AI isn’t the next big thing. It’s the next everything.

We’ve been here before.

Every major technology wave starts as a buzzword and ends as the invisible infrastructure of modern life. I watched it happen with the internet in the ’90s and then again with cloud. Now it’s happening with AI — only faster, deeper, and at enterprise scale.

During my time at IBM Watson, we were building intelligent systems long before “generative” became a headline.  Our algorithms could reason, learn, and assist in decision-making – but they were only as strong as the data underneath. The algorithms were ready; the information architecture wasn’t.

That was the lesson AI fails without great data.

Later at GE Healthcare, I watched AI leave the lab and enter people’s lives. We embedded machine-learning algorithms into medical devices, helping clinicians detect patterns faster to make earlier, more confident decisions. It was one of the first large-scale examples of applied AI (we named it Applied Intelligence) — intelligence engineered into the workflow, not tacked on as an afterthought. The value wasn’t the model, it was the way it quietly elevated the work of real humans,

AI is most powerful when it disappears into the workflow.

From Model Obsession to Infrastructure Thinking

Everyone’s chasing the shiny object – the newest model, the latest tool.  The real magic happens when AI is wired into the business, not layered on top.

At MuleSoft, I learned that integration, not algorithms, is the true unlock.  If your systems can’t talk, your AI can’t learn.  Connectivity isn’t an IT issue – it’s an intelligence strategy.

At OpenText, that principle evolved again.  Information management is the infrastructure.   When your data is connected, classified, governed, and secured, AI moves with trust and context.

The Next Decade Belongs to Integration

We’ve spent two decades building the groundwork — data lakes, APIs, security layers, automation frameworks.  Now it’s time to connect them all.

The winners of this AI decade won’t be the ones with the most pilots. They’ll be the ones with the most interoperability – where data flows freely and securely across every function.

AI isn’t a department anymore.  It’s the new operating system for businesses.

The Real Work Begins Now

Organizations that treat AI as a side project will soon find themselves running two realities: one manual, one intelligent. That’s not sustainable.

If you’re a leader, stop asking, “What can AI do for me?”
Start asking, “How do I re-engineer my business for AI?”

Because AI isn’t the layer on top anymore — it’s the foundation underneath.

Bottom Line

AI won’t replace your business. But the businesses that learn to run on AI will replace the ones that don’t.