- Building and learning
Every System Starts as a Note
Random ideas became notes, useful notes became reusable prompts, and structured prompts became AI workflows. Every step made more of the thinking repeatable.
Thinking Notes began with a simple pattern.
I had ideas everywhere. Some were observations. Some were instructions I wanted to remember. Some were prompts that worked once and might work again. Most began as notes because a note is the fastest way to move a thought out of the mind before it disappears.
The problem was that saving a thought did not make it reusable.
A note could preserve the idea, but the next time I needed it, I still had to reconstruct the reasoning, provide the right inputs, and remember why the approach had worked.
The thought had been captured. The system had not.
A note makes thinking visible
Ideas feel organised while they remain in the mind because the missing context is always nearby.
Writing a note exposes what is actually there.
What is the problem? Which information matters? What decision needs to be made? What should happen next? Which part is reusable, and which part belongs only to this situation?
A good note does not need to answer everything. It creates an object that can be inspected.
Once thinking is visible, it can be improved. Parts can be rearranged. Assumptions can be challenged. A useful line can be separated from the temporary context around it.
This is the first step towards a system: getting the reasoning out of one person's head.
Reusable prompts were more than saved text
As I used AI more, some prompts became worth keeping.
The useful part was not simply the wording. A reusable prompt contained a way of approaching the problem. It described the role of the AI, the information required, the constraints, and the shape of a useful answer.
But a prompt stored as a block of text still asked the next user to know how to operate it. Inputs had to be inserted correctly. Context could be missed. The output could change because one instruction was accidentally removed.
I began to see prompts less as messages and more as small pieces of software.
They had inputs. They contained logic. They produced outputs. They could be tested, versioned, and improved.
That changed the question from “How do I save this prompt?” to “How do I make this way of thinking repeatable?”
From a prompt to a workflow
A workflow gives the thought a structure.
Instead of starting with an empty chat box, a person can be asked for the specific inputs the process needs. The reusable instructions can remain consistent. The result can follow a defined format. What once depended on remembering several steps can become an experience someone else can use.
This is the idea behind the current Thinking Notes product: turning reusable prompts and structured inputs into AI workflows.
The movement was gradual:
Random ideas became notes.
Useful notes became reusable prompts.
Reusable prompts became structured workflows.
The workflows became software.
Each stage externalised more of the thinking.
Externalising thought creates leverage
Keeping a useful method in the mind makes it available only when the person who understands it is present.
Writing it down creates memory. Structuring it creates consistency. Turning it into a workflow creates repeatability. Software can make that workflow available without requiring the original person to guide every use.
This does not mean every thought should become software.
Some notes are valuable precisely because they remain personal and unfinished. Some decisions require judgement that should not be hidden behind a form. Some prompts are used once and never deserve a permanent home.
The useful distinction is repetition.
If a thought keeps returning, if the same inputs are gathered repeatedly, or if the same reasoning has to be reconstructed each time, there may be a system waiting inside the note.
Start by making the thinking inspectable
It is tempting to begin with the application, the interface, or the automation.
But the software can only make a confused process repeat faster.
The first task is to understand the thought itself. What is the real input? Which decisions occur? What makes the output useful? Where does judgement belong? What should happen when the information is incomplete?
Those questions can begin in the smallest possible place.
A note.
Every system I build does not literally begin as a written note. But every useful system begins when the thinking becomes clear enough to examine, share, and repeat.
The note is where that transition often starts.