AI Literacy: Empowering You to Understand AI
Settings and methodology for experienced users of Notebook LM
Examples provided are mainly for researchers and educators

Configure your notebook

You can now configure how this Notebook will respond. Customisation in this setting could result in an output better reflecting your needs. (see the image)

1. Adjust your notebook settings
2. Customise how you wish to use this notebook
3. Choose long (for research) or short.

Identifying and working with sources

1. Finding Expert Sources with NotebookLM
NotebookLM can now actively search the web to find relevant, high-quality sources, rather than relying solely on user-uploaded documents. 
  • Discover Sources Feature: In the sources panel, you can enter a research question or topic, and NotebookLM will search the web for relevant, high-quality sources, including academic papers, reports, and, in some cases, specific, trusted forum posts (like reddit/r/xdadevelopers).
  • Deep Research Mode: Toggle on "Deep Research" for in-depth, AI-powered analysis that scans hundreds of websites to provide a comprehensive report and a list of, for example, 10 highly relevant, vetted sources.
  • Seed Document Research: Start with a high-quality paper, upload it as a PDF, and use the "Discover" tool to find related literature, ensuring your research expands from a reputable foundation.​ 
2. Verifying Credibility within NotebookLM
Once sources are imported, NotebookLM’s AI helps verify the credibility of the information. 
  • Cross-Reference Information: Looking only at the sources in this notebook, identify: 1) Any areas where the sources disagree with each other 2) Any clear contradictions or conflicting claims
  • Check Author/Publisher Authority: Use NotebookLM to scan for institutional affiliations (.edu, .gov) and advanced degrees. Look for peer-reviewed journals, university presses, and respected professional associations (e.g., AMA, IEEE).
  • Assess Currency: Ensure the research is up-to-date by asking NotebookLM to summarize the findings of papers published within the last 5 years.

Step by step process

Introduction 
Notebook LM works best with verified sources. It functions best when sources are specifically focused on a single, specific topic. Be specific  with your research or educational need.

PHASE 1: Deep Research and Source Validation
Use deep search to cover comparative angles with detailed prompt instructions
Step 1
Case study B: Humanities
Context: Sociolinguistic trends.
Prompt: Conduct a deep analysis of the debate around multi-ethnolects in Northern European cities... focus on studies published after 2020 regarding identity construction and language policy.
Step 2
Transform your URLs, PDFs, mp3 audio files, Youtube, etc into structured datasets. Using the Data table feature (in Studio panel)
Prompt:
​I need columns for: investigated sociolinguistic variable, sample size, informant demographics, geographic area, and methodology. Include the source name and key findings regarding linguistic variation. 
Step 3
Validation 

Create a table showing each source with its publication date, author credentials, and whether it's a primary source, secondary analysis, or opinion piece.
PHASE 2: STUDIO CONFIGURATION

​a) Configuring the expert persona
Move beyond generic responses by defining Custom conversational goals with Longer response length.
Persona: The peer reviewer
Senior Professor in sociolinguistics with 20 years of quantitative experience.
Prompt: Act as a 'peer reviewer.' Assess theoretical accuracy (adherece), methodological validity, and statistical significance... Flag methodological weaknesses.
b) Theoretical Synthesis & Mind Mapping
Visualising the interplay between researchers and theoretical frameworks using the Mind Map feature.

Prompt: Visualise connections between concepts... and how they are supported by specific data.
Directed Audio Synthesis
Do not simply click generate. Direct the conversation to expose trade-offs.

Prompt: Focus exclusively on the main disagreements between researchers regarding multi-ethnolects. Explain each perspective clearly and keep the discussion under 15 minutes.
Drafting a report
The Literature Review (Sociolinguistics)

Prompt text:
Write a literature review draft... focus on power relations and linguistic capital. Integrate terms like 'indexicality' and 'prestige' correctly for an expert audience.
The Technical White Paper (Energy)

Prompt text:
Create a technical white paper analysing energy consumption of LLMs vs global streaming traffic. Include methodology comparison, specific data points for kWh per GB, and future sustainability sustainability directions.
Infographics and presentations
Prompt A:
Create a professional infographic comparing across the following three countries. Use a spacious, clean design. Blue pastel colours. 
Prompt B:
Create a 12-slide deck for an international conference... focus on the correlation between social class and dialectal variation.
Flashcards
Prompt:
​Create scenario-based flashcards for field methodology students. Present real dilemmas (e.g., the "Observer's Paradox" or recording ethics). Test the student's ability to choose the right methodological action.

Perplexity to Notebook LM: Example workflow

The process shows how Perplexity AI can be used to gather and analyze information, which is then imported into NotebookLM for organization, synthesis, and presentation. Together, these tools support efficient creation of tables, visual materials, and presentation-ready slides based on the extracted research insights.