I have born witness to the impact of a bad briefs. I have seen military exercises and operations suffer because a lack of detail in the briefing (rare but it happens).
It wasn’t because of poor leadership or bad planning. It happened because the briefing was a complete mess. This led to unwanted outcomes.
I see the exact same thing happening every single day with AI.
People are getting frustrated, blaming the tool for its generic, unhelpful replies. The problem isn’t the AI. The problem is the briefing. We're giving it vague orders and somehow expecting specialist results.
In my 22 years in the Royal Air Force, and the 15 years in the corporate world, I learned one lesson that has stood the test of time: the quality of the outcome is directly tied to the quality of the initial instruction.
You wouldn’t send a soldier on a mission or an engineer to a critical job with a one-line order. You give them context. You provide the bigger picture, you define the rules and constraints, and you make the objective crystal clear.
This is was is called Context Engineering when its applied to AI. It’s the single most important skill for getting professional, reliable results from artificial intelligence. It’s how you turn a generalist tool that knows a little about everything into a specialist assistant that understands your world.
Starting this Monday, I'm dedicating the next seven posts to Context Engineering. I'm going to show you exactly how to apply the principles of a good military briefing to your conversations with AI.
We'll start simply, fixing the common mistakes that hold most people back. Then, we’ll build on that, step-by-step. By the end of the week, you will understand how to construct what I think of as military-grade AI briefings that get you precisely what you need, first time.
Before we begin on Monday, think about the last instruction you gave. Not to an AI, but to a person/s on your team. Was it clear? Did it have enough context for them to succeed without having to guess what you really wanted? How we communicate with people is often mirrored in how we talk to our new digital tools and getting it right changes everything.