Session 3 Slides: The Personal Exocortex

--:-- --
↓ Scroll for more

Session 3

The Personal Exocortex

AI Product Engineering

Block 1: Foundations & The Exocortex

What Is an Exocortex?

An external cognitive system — a digital extension of your memory and reasoning.

  • Your brain: limited working memory, unreliable long-term storage
  • Exocortex: offloads this to a system you design and control
  • For AI engineers: becomes the knowledge layer your AI queries

Static Training Data vs Your Vault

❌ Training Data Only

  • Knowledge cutoff date
  • No domain-specific content
  • Hallucination risk

✅ Your Exocortex

  • Always up to date
  • Your specific domain
  • Grounded, verifiable answers

This vault becomes the foundation of your RAG pipeline.

Why Obsidian?

AI-Ready

Plain Markdown — exactly what AI models work best with

Portable

Copy, search, process with any tool — no lock-in

Durable

Still readable plain text in 20 years

The Four Structural Elements

  1. Atomic Notes — one idea per note, self-contained
  2. Links — [[double brackets]] create a knowledge graph
  3. Tags — #categories that cut across folders
  4. Maps of Content (MOCs) — index notes that navigate the vault

Vault Structure for This Course

vault/
├── 00-Inbox/        # Quick captures
├── 01-Concepts/     # Theory: what things are
├── 02-How-To/       # Practical: how to do things
├── 03-Project/      # Your AI product
│   ├── Product-Spec.md
│   └── Prompts.md
├── 04-References/   # External sources
└── MOC-Course.md    # Master index

Lab: Build Your Vault

  1. Install Obsidian → create vault
  2. Create folder structure
  3. Write three atomic notes from this session
  4. Create MOC-AI-Product.md as project index
  5. Import one external resource as a linked note

Session 3 Summary

  • Exocortex = external knowledge layer your AI will query
  • Obsidian: plain Markdown, portable, AI-ready
  • Four elements: atomic notes, links, tags, MOCs
  • Vault = raw material for RAG (Sessions 8–9)

Next Session: Context Engineering