The Man, the Dog, and the Data Center

Many organizations declare that AI must be part of the product strategy, without examining what that commitment actually entails.

When I was coming up in the world of computing—back in what now feels like the dark ages—we were taught that the data center of the future would be staffed by a man and a dog.

The man was there to feed the dog. The dog was there to bite the man if he touched any of the controls.

At the time, it was a joke. But like many good jokes, it wasn’t really a joke; it was a prediction.

Fast forward to today.

We love AI. We hate data centers.

Everyone wants the benefits of AI—smarter products, automation, insight, speed. But nobody wants the physical reality that makes it possible. Data centers are unpopular. They consume enormous amounts of energy. They require land, water, electricity, infrastructure, and capital. Communities don’t want them nearby, but their local governments promote data centers as job creators, which is, as expected, short-sighted. A few hundred construction jobs for a few months, and then we’re right back to the modern version of the man and the dog.

But here’s the thing: you don’t get AI without data centers. You don’t get the magic without the machinery. And that’s where this stops being a story about product infrastructure and starts being a lesson in product management.

Organizational Competencies 

Here is my definition of strategy in a nutshell:

Products address market needs and problems by leveraging your capabilities in ways competitors do not or will not.

“Address market needs and problems.”

We get ideas from everywhere. You get feedback from your leadership, sales and services teams, as well as the success and support teams. Combine that feedback with your personal experience and domain expertise. And leverage AI analysis to find the patterns.

Every discussion is research—if you write it down and learn from it.

Yet product teams are, by nature, optimistic. And tend to be a little myopic. We focus on building. We often focus more on developing features than on the capabilities needed to support them. We push for better customer outcomes. Better user experiences. Faster workflows. New capabilities that make customers say, “That’s exactly what I needed.”

AI capabilities are the latest and most powerful version of that promise. AI feels like a shortcut to value. Add a model, layer in some intelligence, and suddenly your product looks smarter, more competitive, more modern. Doesn’t look that hard to me!

The key to understanding market needs and problems is real conversations with real customers, not AI prompts and “synthetic users.”

“By leveraging your capabilities.”

Sure, you can add an AI feature to your roadmap in a week. But you can’t add the systems, infrastructure, expertise, and operational discipline quite so easily. Every “intelligent” feature requires a set of organizational competencies that must be developed, integrated, operated, and funded.

This is where a business-oriented approach to product management becomes critical. The role is not to chase the latest trend, but to bring clarity to decisions. To turn good ideas into successful businesses.

That means asking the hard questions.

Do you have the capabilities required to build and sustain this? Does this investment align with how you create value for our customers? What are you giving up in order to pursue this direction? Are you addressing a genuine problem, or simply responding to external pressure?

These are not technical questions; they are business questions. They require a broader perspective than feature-level thinking and a willingness to consider the product as part of a larger system.

What’s striking is how many organizations declare that AI must be part of the product strategy, without examining what that commitment actually entails. AI is the “what” without the “why.”

The conversation tends to focus on keeping up with competitors or meeting perceived market expectations, rather than on understanding the full implications of the decision. Not to mention all the other “number one” priorities already on the roadmap. Which is the most number-one-iest?

Alas, we tend to focus on “can we do it?” rather than “should we?”

The Man and the Dog Never Left

That brings us back to the man and the dog. The original story was not really about staffing levels; it was about control and the evolving relationship between humans and increasingly complex systems. In many ways, that vision has already been realized. The systems have become more automated, more opaque, and more difficult to intervene in directly.

The key difference today is that the man and the dog have not disappeared. They have simply been moved out of sight, into data centers that most of us will never see. We are still dependent on them. We are still paying for them. The only thing that has changed is our visibility into the system.

Sure, we can add AI. But should we? And what can’t we do as a result?

Product management, at its core, is not about chasing what is exciting. There will always be something new that promises to transform the market. The real work lies in understanding what it takes to deliver and support that promise, consistently and at scale, and determining whether the investment is justified.