Reflection

Introduction

In my earlier post, Choosing the Right AI Model, I compared the chat models: ChatGPT, Claude, Grok, Gemini, and the rest. That post is already showing its age, because the conversation has moved. In 2026 the question is less “which model answers best?” and more “which agent should I hand the work to?”

Here’s the difference (super simplified). A model answers. An agent acts. It has its own computer and browser, it signs into your apps, it remembers you, it runs while your laptop is closed, and it comes back when it needs a decision. Nearly every major vendor has put emphasis on shipping new ways to interact and manage agents in the latter half of 2026.

Growing Pains

Imagine you run a gas station that is really successful. People come from far and wide to fill up and to take advantage of your world famous homemade pastries. Everyone agrees that your coffee somehow makes the day just a bit better than the coffee from the service station right across the street. Naturally, you decide to expand - you have been doing everything by yourself but scale requires delegation. So you hire new employees, sketch out the need to add more pumps to the lot and even start looking for locations 2 and 3.

The Governance Paradox

IT departments are frequently buffeted by whiplash projects, thrown back and forth between proactive and reactive, while often balancing capacity. Eventually someone will propose some form of Architecture Review Board (ARB).

The IT goal - prevent the sprawl of technology, get visibility into projects before IT resources are needed or committed, maintain a strong security posture, reduce technical debt and prevent integration problems or redundancies.

Narrow constrained channels are not necessarily a problem. The Strait of Hormuz is a high capacity corridor that provides a far more efficient route for Gulf products to world markets than other realistic alternatives. The same geography that makes the passage strategically sensitive also makes it economically valuable. If it were not the shortest or most useful route, so much trade would not depend on it.

📖 Book Overview

Title: The Model Thinker: What you need to know to make data work for you

Author: Scott E. Page

Published: 2021 (expanded edition)

Genre/Focus: Popular Science

🧠 Why I Read This

As part of my ongoing Doctoral research into Business, Accounting, and Artificial Intelligence, I sought out this book to deepen my understanding of model-driven thinking—a critical skill for modern leadership. Data models are the often-hidden underpinnings of the Information Age, and much of what people refer to as “AI” today is actually fast, layered modeling. My goal was to sharpen my ability to interpret, apply, and challenge models effectively. The Model Thinker delivered valuable tools that I am already applying in strategic decision frameworks.