Why systems die in week three
AI notes and a second brain · Lesson 2 / 22
The same failure script every time
The pattern repeats for everyone. Week one is euphoria: a new app, templates, tags, migrating old records. Week two: plenty of notes, processing falls behind. Week three: opening the system feels unpleasant, because a debt is sitting there. Week four: you are messaging yourself again.
The cause is neither laziness nor the app. The cause is that the system demanded more maintenance than it returned from day one.
Four mechanisms that kill it
- Processing debt. Raw material accumulates faster than you refine it. A hundred unprocessed clippings produce the same feeling as an unread inbox — and the system starts getting avoided.
- Decisions at capture time. If saving requires choosing a folder, a tag and a note type, you will not save. Every decision adds seconds and a chance of skipping.
- No feedback. The system returns nothing: you put things in and never take anything out. A habit without reward does not survive.
- Copying somebody else's system. Schemes designed by people who write about notes professionally assume volumes you do not have.
Survival test
The system will outlive a month if: — capture takes under 5 seconds and requires no choice; — processing works in 10-minute chunks, not "when I get a free evening"; — skipping a week breaks nothing and creates no debt; — you can name at least one scenario where you take from it rather than add to it. Three yeses out of four means it lives.
Worth admitting separately: an abandoned system is not a disaster. The damage starts when you launch a fresh one from scratch in a different app and run the same cycle twice a year.
Cheat sheet
- Systems die of processing debt, not missing features.
- Any decision at capture time lowers the odds you record anything.
- Habits survive on return: you have to take things out.
- Skipping a week must break nothing.
Recall your last abandoned note system — an app, a notebook, a cloud folder. Describe when you started, when you quit, roughly how many entries you made, and what the last entry was. Diagnose which of the four mechanisms from the lesson actually killed it and give concrete evidence, for example the number of unprocessed items at the moment you stopped. Then run that system through the survival test and state which items it failed.