---
type: Article
title: Build Your Own Regenerative AI Partner (The NotebookLM Primer)
slug: notebook-primer
url: https://shannondobbs.com/notebook-primer/
resource: https://shannondobbs.com/notebook-primer/
surface: shannondobbs.com
register: implementation-guide
audience_pain_vectors:
- ai-curious-mission-worker
- impact-org-no-bandwidth
- accessibility-constrained-operator
- community-food-organizer
- resource-constrained-household
domains:
- ai
- coordination
- water
- energy
- community
thesis_tags:
- coordination-as-infrastructure
- structural-vacuum
- fractal-snapshot
receipts:
- Google NotebookLM, free, 3 URLs + a Google account
- White Beard Strategies (Jonathan Mast) - disclosed affiliate
- Wispr Flow $15/mo - disclosed affiliate
- context-loading cuts AI cycles from 10-20 to 2-3, reducing data-center water/energy
  draw
- 'RIA / Sori Village pilot: offline-capable AI on local hardware for low-bandwidth
  bioregions'
- CC BY 4.0
key_reframe: The substrate matters more than the prompt. Self-contextualization fluency
  opens AI access while reducing AI's environmental footprint at the same time --
  most of the AI-and-climate conversation treats those as a tradeoff. They aren't.
whimsy_marker: An AI workspace without good material is just a fancy notepad.
thread_anchor: https://shannondobbs.com/the-thread/
description: The substrate matters more than the prompt. Self-contextualization fluency
  opens AI access while reducing AI's environmental footprint at the same time --
  most of the AI-and-climate conversation treats those as a tradeoff. They aren't.
tags:
- coordination-as-infrastructure
- structural-vacuum
- fractal-snapshot
- ai
- coordination
- water
- energy
- community
---

# Build Your Own Regenerative AI Partner in 15 Minutes
*A free three-step NotebookLM walkthrough. No technical background, no subscription — three URLs and a Google account. (Affiliate disclosure: page 2 has disclosed affiliate links to tools we actually use; they help fund the pilots.)*

What you're really learning is bigger than NotebookLM: how to load AI with the context that makes it useful for *your* work — a meta-skill that opens almost every published resource on the internet.

**Step 1 — Open your workspace.** Most people doing regenerative work are drowning in information they can't use. Google quietly released NotebookLM (free, any Google account, nothing to install). Create a new notebook. An AI workspace without good material is just a fancy notepad — that's the next problem.

**Step 2 — Feed it the right substrate.** When you give an AI good source material, the same tool that gave you generic nonsense yesterday gives sharp strategic analysis today. Load three pre-built sources: *Below the Radar* (UN-facing regenerative framing), *The Regenerative Strategic Partner Gem* (the core orientation document — start here), and optionally the *Fire Defense Field Guide* (fire/drought regions). The substrate matters more than the prompt.

**Step 3 — Put it to work.** This is the same strategic capacity that costs other organizations thousands a month — free, yours. The leverage isn't searching, it's *thinking with substrate*: not "what is regenerative agriculture" (that's a Google search) but "based on these documents, how would I make a case to [this funder] for [this project]?" Within a few weeks it knows your context better than most consultants you could hire.

**Going deeper (page 2):** disclosed affiliate recommendations — Jonathan Mast / White Beard Strategies (my AI mentor; the AI Bazaar ecosystem) and Wispr Flow (voice-to-text, $15/mo, a real help for field work and chronic conditions that make typing hard).

**Why it matters — AI fluency is climate infrastructure.** Every AI iteration consumes data-center water and energy, often in the same drought-prone bioregions where regenerative work operates. Context-loaded AI gets to a conclusion in 2-3 cycles instead of 10-20. Fewer iterations, less compute, less pressure on watersheds. And not everyone has the bandwidth to use this yet — operators in places like Sori Village do the same work without dependable grids, which is why RIA's pilot is building offline-capable AI on local hardware. *(CC BY 4.0 — share it, fork it for your bioregion.)*