Knowledge compounds when it's linked, not just stored

Saving things feels like progress, but a folder of clippings does nothing on its own. You can hoard a thousand highlights and still arrive at a blank page, because storage isn't recall — the notes sit there, unconnected, and none of them points you anywhere. The value was never in having the idea filed; it was in being able to reach it when a different idea needed it.

Links are what turn a pile into a web. When one note names another, each new connection makes both ends more findable, and the whole collection grows in usefulness faster than it grows in size. That's why Evergreen notes are written to evolve, not to file — they're written to be linked — and why Structure beats freeform text for working with AI when you want to actually work with what you saved. Even a single page can hold the threads for you, as A page can remember what you learn.

This is also how retrieval-augmented AI earns its keep. It doesn't conjure knowledge; it retrieves what you already connected, and the better your connections, the better what comes back. See the frameworks for memory systems.

Next: write notes that improve over time, since Evergreen notes are written to evolve, not to file.