Daniel Lyons' Notes

Learn 80% of NotebookLM in Under 13 Minutes!

Description

Learn how to use Google's NotebookLM to grow your business.

Notes

00:00 When to Use Notebook LM

  • Three key criteria for Notebook LM performance:
    • Very low tolerance for AI hallucinations
    • Working with information across different locations, formats, and mediums (documents, slides, text, video, audio)
    • Need to transform fragmented information into cohesive output

00:37 Initial Setup and Navigation

  • Press Escape when overwhelmed by new notebook interface
  • Bookmark homepage for easy reference
  • Change to list view and sort by title for organization

01:00 Health Reports Notebook Example

01:00 Creating and Uploading Sources

  • Upload multiple annual health checkup reports
  • Add PDFs and YouTube videos as sources
  • Example sources: Health reports, "Tools of Titans" by Tim Ferris, videos on uric acid and fasting

01:41 Source Guide and Chat Interface

  • Notebook LM generates concise summaries of dense medical reports
  • Clicking key topics expands them using ALL selected sources, not just one
  • Deselect sources to exclude them from analysis

03:06 Practical Health Analysis Tasks

  • Identify top 10 health trends across multiple reports
  • Cross-reference health issues with video sources (e.g., uric acid levels with video on the topic)
  • Review inline citations to verify information sources
  • 03:48 Important: Save outputs to notes or they disappear on reload
  • Ask contextual questions combining multiple sources (e.g., "Should I fast given my health reports?")

04:56 Key Difference: Hallucination vs Creativity

  • Notebook LM is fine-tuned to hallucinate less
  • Google Gemini is optimized for speed and creativity but more prone to hallucination
  • Multiple saved notes can be converted to a standalone source

06:01 Use Case 1: Focused Knowledge Retrieval

06:05 Equipment Manuals

  • Upload all product user manuals in one notebook
  • Ask specific questions like firmware updates or camera settings
  • Pro tip: Find PDFs by searching "[product name] user manual PDF"
  • Alternative: Add website directly as source (bypass blocks by copy-pasting text)

07:00 Business and HR Applications

  • Tax and accounting notebook: Add government tax codes and audit reports
  • Ask about tax obligations, financial trends, eligibility questions
  • Recruiting notebook: Add HR guidelines, performance rubrics, candidate resumes, interview notes
  • Pro tip: Deselect documents from other candidates to avoid cross-contamination

08:10 Use Case 2: Project Context Engine

  • Create one notebook per project with meeting notes and project plans
  • Particularly valuable for project and program managers who work with scattered information
  • 08:41 Generate deliverables:
    • High-level briefing documents for senior leaders
    • Campaign timelines for key milestones
    • FAQ documents for colleagues
  • 09:04 Pro tip: Upload Zoom/Google Meet transcripts for:
    • Outstanding tasks identification
    • Meeting recap emails
    • Strategy and learnings from previous campaigns
  • Pro tip: Start with suggested questions if struggling to get started

10:00 Use Case 3: Earnings Analysis and AI Strategy Research

  • Create earnings analysis notebook with company reports and analyst articles
  • Ask targeted questions: "What is Google's monetization strategy with AI?"
  • Ask broader comparative questions across companies
  • 10:48 Generate customized audio overviews (podcasts):
    • Click customize audio overview
    • Provide specific instructions (focus areas, audience level)
    • Listen during commutes
  • 11:23 Pro tip: When using Google Docs/Slides as sources, click to resync file after changes

11:36 Final Thoughts and Recommendations

  • Notebook LM outputs are reliable but not optimized for creativity—use Gemini or Claude for final polish
  • Massive capacity: ~25 million words per notebook (vs. 500K for Gemini, 100K for Claude, 64K for ChatGPT)
  • Workaround: Combine multiple documents into one file despite 20-source limit
  • Source quality is critical—use well-established publications, not clickbait blogs
Learn 80% of NotebookLM in Under 13 Minutes!
Interactive graph
On this page
Description
Notes
00:00 When to Use Notebook LM
00:37 Initial Setup and Navigation
01:00 Health Reports Notebook Example
01:00 Creating and Uploading Sources
01:41 Source Guide and Chat Interface
03:06 Practical Health Analysis Tasks
04:56 Key Difference: Hallucination vs Creativity
06:01 Use Case 1: Focused Knowledge Retrieval
06:05 Equipment Manuals
07:00 Business and HR Applications
08:10 Use Case 2: Project Context Engine
10:00 Use Case 3: Earnings Analysis and AI Strategy Research
11:36 Final Thoughts and Recommendations