Building a Note System: The "Second Brain" Approach

We have all lived the same experience: "I read/wrote this somewhere, but I can't find it". The thoughts lie scattered across the phone notes, the screenshots, the chats, the notebooks; when needed, none is at hand. The note taking system is that loss's solution, and its fashionable name is "the second brain": move the memory from the brain into the system, let the brain busy itself with thinking. It is not as esoteric as it sounds; it is three-step engineering.

This article gives that engineering: the capture-process-use loop, the practical rules, the tool choice and the new layer AI brings.

The loop: capture, process, use

The working note system's skeleton is three steps. The capture: every thought-and-fact lands in one channel (the single-inbox rule: one app/section on the phone; the five different places called "I'll systematise later" are systemlessness itself); the friction must be minimal: a note that cannot be written in two taps does not get written. The processing: 5 minutes daily/every other day: the inbox empties, every note goes to its address (if a task, to the task system; if knowledge, to the note base; if rubbish, deleted). The use: the notes exist not to be written but to be returned to: opening the relevant notes when starting a project, the learning repetitions, the material for writing-and-decisions. Where the systems die is usually the second step: things get captured, not processed, and the archive turns into a dump; the 5-minute ritual is the system's heartbeat.

The processing rules: a valuable note's anatomy

The practical answer to the folder-versus-tag debate: in the search era the deep folder trees have aged; a few wide categories + good titles + the search habit suffice. The system's simplicity is its endurance's condition.

The tool choice: three philosophies

The note tools split into three families: the simple-linear (the Apple Notes/Google Keep class: the instant-capture advantage, the structure weakness; fully legitimate for the start), the database-based (the Notion class: structured, linked, team-friendly; the standard in business use) and the link-based (the Obsidian class: the links between notes, the local files, the thought-map philosophy; the deep knowledge workers love it). The choice rule from the context: the one writing mostly from the phone should start in the simple family, the one working with a team look at the database family, the research-and-writing person at the link family. And the migration moratorium here too: changing the tool is not renewing the system — it is usually fleeing it.

The AI layer: the notes talking

The new era's contribution: the note archive is no longer passive. The possibilities: the conversation with your own notes (by the RAG logic: "what notes do I have about this client?", "what conclusion did I reach on this topic last year?"; the Notion-AI-type built-in features or handing the notes to an AI tool), the automatic structuring (the voice note → the text → the title-and-tag proposal), the summary bridges (the two-sentence own-words note draft from a long article; the final edit yours) and the periodic synthesis (the "recurring themes" report from the month's notes: your own thinking's mirror). With one warning: AI does not replace the processing ritual; a garbage archive's summary is a garbage summary. The loop (capture-process-use) stays; AI is its accelerator.

Sources and further reading

Where to verify the source

The second brain approach's source:

• Building a Second Brain: the method's official resource

Continuing the topic

This line's neighbouring articles:

• The database-based tool
• The system architecture
• The link with learning
• The team version
• Other articles on this topic

One decision today: choose your single inbox and pull it onto the phone's home screen. From tomorrow let every thought land there and get processed in the evening's 5 minutes. The second brain is born not of a grand structure but of this small loop.