Keeping what I learn
A two-layer knowledge system — a workshop for doing the work, a library for what lasts — and a memory that survives between sessions.
I separated the place where work happens from the place where durable knowledge lives, with a deliberate promotion step between them. The judgement is in the boundary: most notes are disposable, a few are permanent, and conflating the two is how knowledge systems rot.
The default failure of any personal knowledge system is that everything goes in the same pile — fast notes, half-thoughts, and the few ideas actually worth keeping — and the signal drowns. The fix isn't a better note-taking app. It's a boundary.
So I run two layers, deliberately kept apart.
Workshop and library
- The workshop is where the work happens — daily notes, decisions, project material, in-flight thinking. It's time-bound by design. Most of what lands here is meant to be disposable: useful this week, irrelevant next month, and that's fine.
- The library is a separate store of durable concepts — the things that stay true across projects. It's small on purpose. Different repository, different rules, navigable in a tool built for connected notes rather than chronological ones.
Between them sits a promotion step: a weekly review where I ask what, from the week's workshop churn, has earned a place in the library. Capture is wide; promotion is narrow. Capturing something creates a memory, not a commitment — and keeping that distinction is what stops the durable layer filling up with noise.
Why the boundary is the point
It would be simpler to keep everything in one place and trust search to sort it out. I chose the boundary on purpose, because the two layers answer different questions and decay at different rates. The workshop is allowed to be messy and ephemeral; the library has to stay trustworthy. Mixing them means either the durable knowledge gets buried in transient noise, or the transient notes get over-curated as if they were permanent. Both are failure modes. The separation is what lets each layer be good at its one job.
Memory that survives the session
There's a third piece that complements the two layers: a small, file-based memory that persists across working sessions — durable facts about how I work, recalled before anything substantive happens. It's a different job from the library (operational memory rather than durable concepts), so it gets its own write-up: giving AI a memory that lasts.
The whole thing is sanitized by design: what's described here is the architecture and the rules. The contents — the actual notes, the personal material — stay where they belong, behind the boundary. That restraint is the same one that runs through the rest of the work: the structure is shareable; the contents are not.