Technology cycles move faster than the lessons they teach.

I've spent multiple decades watching organizations adopt systems that promise transformation—containers, cloud platforms, distributed architectures, now autonomous agents. The tools change. The patterns don't.

What follows are not predictions or case studies. They're retrospectives—re-examining what I wrote years ago about where certain technologies would lead, measured against where they actually went.

Not to prove I was right. To show what patterns hold across cycles.

The Framework

How I Evaluate Technology Cycles
A framework built from watching infrastructure evolve across thirty years—from physical servers to autonomous agents.

The Retrospectives

Cloud Lock-In: 2019 → 2025
Five years ago, I wrote about why "cloud native" would become the next vendor lock-in cycle, not an escape from it. Here's what happened.
Containers to the Edge: 2018 → 2025
In 2018, I argued that container orchestration would fragment, not consolidate. Seven years later, here's where that thesis landed.

Why This Matters

If you're evaluating AI agents, governance platforms, or any technology promising to "change everything"—these patterns apply.

The architectures change.
The adoption curves don't.
The governance problems certainly don't.

This is how I think about systems. It's why Equilateral exists the way it does.