Chapter takeaways that compound.

The pain

• —You finish a book, write 5 takeaways, close the file. Six months later you can't recall whether it was Munger or Buffett who said the thing about inversion.
• —Online courses scatter notes across Notion / Apple Notes / a half-finished Obsidian vault. None of them deploy to your AI.
• —Each chapter feels isolated when you write it. The connections to other books / courses don't surface until you're already trying to remember them.
• —Re-reading is expensive. Without compounding notes, every revisit starts from page 1.

What you do in Memory.Wiki

1. 1
   One doc per chapter (or per lecture)

   Drop in your raw takeaways. Memory.Wiki doesn't care if it's bullet form or prose. The title is the chapter — that's the H1, that's the address.
2. 2
   Let recurring concepts surface themselves

   Concept index auto-runs. After 3-5 chapters, the cross-chapter concepts (inversion, second-order thinking, etc.) get the orange dot in the sidebar Concepts list.
3. 3
   Bundle by theme — not by book

   "Mental models for risk" might pull 3 chapters from one book + 2 from another + a course module. The Bundle is the curriculum you wish someone had assembled.
4. 4
   Ask the hub what you've learned

   "What have I read about inversion?" Hub recall returns every chapter that touched the concept with the exact passages. Better than your own memory.

What you get back

• ✓Concept clusters emerge as you read more — the wiki compounds without manual cross-linking.
• ✓Compiling a Brief from a theme-bundle gives you the synthesis you'd otherwise have to write yourself.
• ✓Year-end "what did I learn this year?" answers itself.
• ✓When you write something new, paste the hub URL into Claude — it argues with your past self using your own notes.

Worked example

Try it with what's on your desk right now.

No signup. Drop in your first doc and the URL is yours.

Other shapes the URL takes