# AI Wiki - PKM - Atomic Notes An atomic note captures exactly one idea, concept, or insight in its own standalone note. The term draws from physics: just as atoms are the fundamental building blocks of matter, atomic notes are the fundamental building blocks of a knowledge system. ## Why Atomicity Matters Atomic notes are not merely a note-taking technique. They are an architectural choice for building a scalable, maintainable Personal Knowledge Graph. By breaking knowledge into its most essential components, atomic notes enable three critical properties: **Composability.** Because each atomic note captures a single idea in isolation, it can be freely recombined with other notes. This composability is their key property. Ideas locked inside long, multi-topic notes cannot be independently referenced, linked, or reused. Atomic notes free up ideas, making them available for unexpected combinations and emergent insights. **Connectability.** An atomic note has a clear identity, which makes it easy to link to and from. Connectability is the primary atomicity heuristic: the right level of granularity is determined by how well a note can connect to other notes. If a note tries to say too many things, it becomes hard to link precisely. If it says too little, it has no substance to connect. **Findability.** A note about one thing is easier to title, tag, and search for than a note that covers five topics. When you need to find "that idea about composability," you go straight to the note titled exactly that. ## How to Write Better Atomic Notes Six questions to ask yourself when creating an atomic note: 1. **What can I do with this idea?** — Tests actionability 2. **What attracts me to this idea?** — Surfaces personal relevance 3. **What other ideas does this one validate?** — Identifies supporting connections 4. **What other ideas does this contradict?** — Identifies tension and productive disagreement 5. **Could this idea change by taking a different point of view?** — Tests robustness 6. **How does this note explain other ones in my system?** — Tests integration value These questions shift the focus from "did I capture this correctly?" to "how does this idea live in my knowledge system?" ## Implementing Atomic Notes **One idea per note.** This is the non-negotiable rule. If you find yourself writing "also..." or "another thing..." in a note, split it. **Write in your own words.** Paraphrasing forces understanding. A copy-pasted excerpt is not an atomic note; it is a highlight waiting to be processed. **Title as claim or concept.** The title should communicate the note's core idea. "Atomic notes free up ideas" is better than "Notes on atomicity." A good title functions as a one-line summary. **Link generously.** Every atomic note should link to at least 2-3 related notes. These links are how individual atoms form molecules of understanding. **Let notes evolve.** An atomic note is not frozen at creation. As your understanding deepens, refine the note. Add nuance, correct errors, strengthen the argument. ## Relationship to the Zettelkasten The concept of atomic notes is deeply connected to the [[AI Wiki - PKM - Zettelkasten Method]], where Niklas Luhmann's permanent notes (Zettels) were inherently atomic. Each slip of paper in his physical system captured one thought, enabling the dense web of cross-references that powered his prolific output. In the [[AI Wiki - PKM - Note-Taking Taxonomy]], atomic notes correspond most closely to permanent notes: ideas you have processed, rewritten in your own words, and integrated into your knowledge graph. ## Common Mistakes - **Too large:** Multi-paragraph notes covering several sub-topics. Split them. - **Too small:** A single sentence with no context or substance. An atomic note should be self-contained enough to be understood on its own. - **Copy-paste without processing:** Highlights and quotes are inputs, not atomic notes. Process them first. - **No links:** An unlinked atomic note is an orphan. It has no way to compound in value. ## Key Points - One idea per note; composability and connectability are the defining properties - Connectability is the heuristic for right-sizing atomicity - Write in your own words, title as a claim, link generously - Atomic notes are an architectural choice, not just a technique ## Open Questions - Is there a practical upper limit on the number of atomic notes a system can sustain before noise overwhelms signal? - How should AI tools assist in atomizing long-form captures? ## References - Vault: Atomic notes, Benefits of atomic notes, How to implement atomic notes in your PKM system, How to create better atomic notes, Atomic notes are more than just a note-taking technique, Atomic notes free up ideas - Sönke Ahrens, "How to Take Smart Notes" (2017) - Andy Matuschak, "Evergreen notes should be atomic" (working notes) ## Related - [[AI Wiki - PKM - Zettelkasten Method]] - [[AI Wiki - PKM - Evergreen Notes]] - [[AI Wiki - PKM - Note-Taking Taxonomy]] - [[AI Wiki - PKM - Compounding Knowledge]] - [[AI Wiki - PKM - The Capture Habit]] - [[AI Wiki - PKM - Writing as Thinking]] - [[AI Wiki - PKM - Personal Knowledge Graphs]]