In an AI World Full of Noise, I’m Being Skeptical on Purpose

I’ve been labeled skeptical a bit recently around AI. I’m fine with that.

But it’s worth saying what that skepticism actually is (and what it isn’t).

I’m not anti‑innovation. I build things for a living. I like new ideas. I like progress. I like when technology actually moves the ball forward. What I’m against is noise. And right now, MSPs are drowning in it.

Every cycle brings a new framework, a new model, a new set of tools, a new abstraction layer that promises to “change everything.” The language is confident. The diagrams are clean. The demos are impressive. And yet, when you step back, a lot of it doesn’t survive contact with reality.

That’s where my skepticism comes from. I’ve been in this space for 15+ years. I know what the beginning of a cycle looks like. It often follows the same patterns and the same experimentation. And for a decade and a half, it’s mostly landed in the same ending position, whether cloud, cyber, or AI: repeatable and monetizable at scale.

I don’t start by believing vendors. I don’t start by assuming the abstraction is necessary. I don’t start by trusting that because something is popular, it’s useful. I start by asking a much more boring question: Does this actually hold up when you try to run it, scale it, support it, and charge money for it?

Most things don’t fail because they’re bad ideas. They fail because they’re fragile, expensive, hard to explain, or impossible to operationalize. Or they only work under perfect conditions that never exist outside a demo. And in today’s era, with the break neck speed that things get done, none of that is acceptable.

I think a lot of people miss this. Especially in AI.

There’s a temptation to treat intelligence as magic instead of infrastructure. To stack more layers, more orchestration, more cleverness on top, and assume value will appear. But intelligence that can’t be repeated, governed, or monetized isn’t progress. It’s a science project, and its irresponsible if you’re doing it in your customers environments.

Healthy skepticism is how you protect yourself from that.

It forces you to slow down and separate what’s interesting from what’s durable. What sounds smart from what actually compounds. What helps one team ship a demo from what helps an organization operate at scale. And it’s the exercise you need to stop feeling so overwhelmed with the noise.

This is where my head goes with it: if something can’t be explained simply, deployed repeatdely, and improved incrementally, it’s probably not ready.

So yes, I’m skeptical. On purpose. Because skepticism is how you calm the noise. And once the noise is gone, the real work can start.

Look forward to sharing more of my learnings soon! (They’ll be a little less skeptical, I promise)

Newer Post >