The more I learn about systems thinking and game theory, the more I see that context is as important, if not more important than the actual thing one wants to achieve.
There’s a difference between topic-thinking and terrain-thinking.
Let’s say you want to solve a problem or achieve something specific. Topic-thinking will compel you to look into the mechanics of how to get from A to B. This approach is understandable because it’s linear. You can easily break it into procedural steps. And once you get a handle on it, you can even automate those steps using LLM skills or agents. Yes, you might run some research to understand the conditions in which you operate or the characteristics of the players you’re engaging with, but this is only a surface-level action that simply allows you to tweak those steps accordingly.
Topic-thinking becomes even more pervasive if we’re talking about great professionals who work based on methodology-as-a-substitute-for-seeing. They already have a personal or borrowed playbook that can be easily adapted to different situations. Even when complex, those are still often bounded domains. You know the saying: everything looks like a nail when you have a hammer.
The problem is that this type of thinking might not survive contact with reality. You might apply well-known playbooks or focus specifically on the mechanics around the topic of your thinking, but this doesn’t guarantee an optimal result. Topic-thinking feels rigorous because it has steps. But rigor applied to the wrong object is just sophisticated misdirection.
A company decides to use AI to increase team productivity. They identify repetitive tasks, build automation workflows, measure output per person. Productivity metrics go up. But gradually, work quality drops and coordination breaks down. No one mapped how those “repetitive” tasks were actually the moments where people built shared context: the handoffs, the reviews, the quick back-and-forth that kept everyone aligned. They optimized the visible mechanics and destroyed the invisible infrastructure.
Terrain-thinking, though, is an entirely different dragon. This type of thinking focuses on mapping and understanding the structure around the problem you want to solve or the goal you want to achieve. Its main discipline revolves around the underlying logic where you’re about to act:
A different company asks: before we automate anything, how does good work actually happen here? They discover that productivity isn’t bottlenecked by slow tasks, but by how context moves between people. The slow parts aren’t waste. Some of them are load-bearing. So they deploy AI only where removing a task doesn’t sever a knowledge flow, and protect the parts that look inefficient but are actually holding the operation together.
Terrain-thinking asks whether the topic even survives contact with the actual structure around it. Topic-thinking can be executed perfectly and still be wrong, because it never questioned whether it was pointed at the right thing. Terrain-thinking might tell you to abandon the topic entirely, or reshape it, or realize the real leverage is somewhere adjacent that you couldn’t see while staring at the topic itself.
The next time you’re about to solve a problem or achieve a goal, stop. You’re probably already looking at the topic. Turn around. Look at everything around it. The structure you see will likely change what you do.
