The Black Box
You think constantly. Thousands of decisions per day. What to eat, what to say, which email to answer first, whether to take that job offer, how to solve that problem at work. Yet if I asked you: “How do you think?” you’d probably struggle to answer. Most people say things like, “I just… think about it.” or “It depends on the situation.”
Thinking feels mysterious. Like a black box you put problems into and answers come out. This mystique creates problems:
Problem 1: You can’t improve what you can’t see.
If thinking is an invisible, intuitive process, how do you get better at it? Read more? Think harder? Hope for divine inspiration? The advice is vague because the process is invisible.
Problem 2: You can’t teach what you can’t articulate.
Expert decisions look like magic to novices. “How did you know to do that?” → “Experience.” But what is experience? If you can’t explain your thinking process, you can’t transfer expertise. The expert retires, the expertise leaves with them.
Problem 3: You can’t debug failures.
When a decision goes wrong, what do you fix? “I should have thought more carefully” isn’t actionable. Without seeing the process, you can’t identify which step failed.
Problem 4: You can’t automate or scale.
If thinking is a mysterious human intuition, it’s stuck in human heads. You can’t build systems, structure AI processes, or create repeatable organizational capabilities. Everything depends on individual genius.
The Revelation
But what if thinking isn’t mysterious at all?
What if, when you carefully observe, thinking follows clear, repeatable processes —just like walking, typing, or driving? You learned to walk by executing a process (shift weight, move leg, balance, repeat). You learned to type by executing a process (finger positioning, muscle memory, visual feedback). You learned to drive by executing a process (mirror, signal, accelerate, steer).
None of these feel like “processes” anymore. They feel automatic, intuitive, and natural. But they’re still processes. You’ve just internalized them through practice.
Hypothesis: Thinking is the same. You execute clear processes constantly, but you’ve practiced them so much you no longer notice the steps.
The Journey
We’ll explore thinking at three levels of complexity:
Level 1: Simple Decisions
We’ll start with something mundane: deciding what to eat for lunch.
Seems simple, right? You just… choose. But when we slow down and observe, we’ll discover a clear 5-step process:
- Recognize the need
- Verify the context
- Enumerate options
- Evaluate constraints
- Select best fit
You run this process unconsciously. But it’s still a process. It is structured, repeatable, and can be improved by isolating the individual steps.
Level 2: Goal Seeking / Path Finding
How do you get from where you are (A) to where you want to be (B)? How do I write and publish a book? How should we scale our business 10×? These feel fundamentally different from “what’s for lunch.” But we’ll discover they follow a similar process pattern , just with more depth:
- Assess current situation (A)
- Define desired goal (B)
- Decompose into possible paths
- Evaluate trade-offs of each path
- Identify risks and unknowns
- Design mitigation strategies
In the above thought process, step #3 is particularly important and complex. Decomposition is the root thought process behind strategic thinking . We will expand on this concept later on.
Level 3: Complex Reasoning
Are thought process patterns unique to humans, or universal to effective thinking itself?
To find out, we’ll examine how AI systems reason:
- Chain-of-Thought (CoT): Decompose problem → Solve sub-problems → Aggregate results
- Tree-of-Thought (ToT): Enumerate options → Evaluate each → Explore promising → Prune failures
- O1-style reasoning: Internal A→B framework (assess, decompose, explore, backtrack, validate)
The revelation: AI reasoning systems rediscover the same process patterns we see in human decision-making. Not because AI mimics humans. Because effective thinking requires process structure —whether the thinker is human or machine.
The Universal Pattern
Once we’ve seen thinking processes at all three levels, we’ll synthesize the elements that appear in ALL effective thinking.
Spoiler: 1. Problem decomposition 2. Option generation 3. Structured evaluation 4. Constraint awareness 5. Risk/failure detection 6. Iterative refinement. These elements appear whether you’re choosing lunch, planning a career change, or designing an AI reasoning system. This is not a coincidence. This is fundamental architecture.
Thinking is process execution. Better processes → better thinking.
Why This Matters
Understanding thinking as process has immediate practical implications:
For individuals: If you can make your thought process traceable , you can analyze and improve it. You can also document your thought process, which means you can share expertise, not just conclusions.
For teams: You can replace emotional contentious arguments with logical and objective differential analysis. If Person A and Person B disagree about something, that disagreement can only be rooted in two possible places: evidence or mental models . Or put another way, what does a person believe is most likely true (critical thinking) and how do they process that belief (logic and heuristics).
For AI systems: You can design reasoning with explicit processes. CoT/ToT prove this works. You can align AI expertise with human expertise by encoding proven decision-making frameworks and heuristics. You can make AI decision making more explainable by examining the inputs and outputs of individual reasoning steps.
For organizations: You can build a learning organization by converting expert intuition into repeatable processes. This allows you to scale expertise. It enables training and automation. And most importantly, it makes continuous improvement systematic and achievable.
Intuition and Reflexive Decision-Making
“Some of my best decisions come from intuition, not a mechanical process!”
Fair point. Let’s address it upfront.
Intuition is not the absence of process. It is a manifestation of neural distillation and compression through pattern matching.
When an expert chess player “just knows” the right move, they’re not using magic. They’re pattern-matching against 10,000+ hours of experience. They have many times gone through the thought process of “evaluate board state → enumerate candidate moves → assess threats and opportunities → select best move”. This results in a distillation or compression of input → decision into a single reflexive action.
Your intuition and reflexive decision-making are great for cases where stakes are low, poor decisions are easily reversed, or time is short. But diagnosing bad decisions that are based on reflexive thinking is almost impossible. If you try to explain to yourself or others how you made that decision, you will end up rationalizing and back-filling the structured thought process. But you did not explicitly follow that thought process.
Next in the Series:
Part 2: The lunch decision Part 3: A→B goal achievement, how decomposition leads to strategy Part 4: AI reasoning (CoT, ToT, O1)