By Michael Spencer and Alex McFarland from AI Supremacy

I have started a new Newsletter on the AI chip sector called Semiconductor Things™.

• This Newsletter was built to solve the pain point of getting the latest news on A.I. chips, semiconductors, datacenter innovation and chip news.
• If that interests you, you’re welcome to sign-up in its early form.

Learn more.

Turn messy community discussions into clear insights with this technique.

Share

In September, 2024 Audio Overviews went viral creating a moment in time when Google Labs’ NotebookLM had its own “ChatGPT” like moment. That Generative AI could magically turn any topic into a synthetic general purpose podcast felt weirdly incredible! 

That’s when Alex McFarland of AI Disruptor wrote one of the most shared guest posts on this publication ever: How to use NotebookLM for personalized knowledge synthesis. It’s been shared 716 times so far and has resulted in thousands of clicks to Learn About and Google’s latest AI Tools like Deep Research. 

Since NotebookLM keeps improving now with a premium version, people have been using it in creative and diverse ways in an increasingly personalized manner customized to their research needs and preferences. This guest post is the 2nd in our series on NotebookLM by Alex McFarland. 

Try Deep Research

What is NotebookLM? 

If you haven’t heard about NotebookLM, let me give you a quick definition and with Google Gemini 2’s evolution this isn’t final eiter – the tool is constantly being improved and upgraded: definition: 

NotebookLM is an advanced AI-powered tool developed by Google, designed to enhance the way users engage with and make sense of their information. It serves as a virtual research assistant that utilizes the capabilities of language models to assist users in summarizing, organizing, and analyzing their notes and documents.

“Think Smarter, Not Harder.” 

How you can now customize Audio Overviews is also pretty amazing. From Project Tailwind to the beast we see today, NotebookLM’s story is fairly fascinating as an AI experimental product now reaching the mainstream and millions of users. It’s just what Google and Google Deepmind needed heading into 2025. 

Try NotebookLM

• One of NotebookLM’s creators is also here on this platform of the Newsletter Adjacent Possible, Steven Johnson. 
• Since Alex’s first guide on NotebookLM that went viral in November, his Newsletter has also evolved with many other useful guides and videos, that I want to highlight: (he’s also thinking about starting a Portuguese version for my readers in  Brazil, you know who you are). 

Posts by Alex McFarland 

1. How to digest big tech events like CES with NotebookLM
2. How to setup and use ChatGPT’s new “Tasks” 
3. Here’s what nobody is telling you about AI agents in 2025

I think NotebookLM Plus is a big deal and the ability to jump in and join the formerly annoyed AIs is fascinating. The “Interactive mode” of Audio Overviews can be a lot of fun. 

The December, 2024 redesign also means NotebookLM keeps improving. The redesign organizes NotebookLM into three areas. The “Sources” panel manages all the information that’s central to your project. The “Chat” panel lets you discuss your sources through a conversational AI interface with citations. The “Studio” panel lets you create new things from your sources with a single click, like Study Guides, Briefing Docs and Audio Overviews.

• Sources
• Chat
• Studio 

Share AI Supremacy 

In combination with Google tools like Learn About, Deep Research and other tools it’s using Google as a research assistant is getting legit in 2025. 

NotebookLM’s going viral was affirmation that Google’s AI tools were getting more interesting to consumers. Fast forward to today and NotebookLM even has its own dedicated Twitter/X account. Let’s move on to the guide now. 

By Alex McFarland, January, 2025. 

Read Part 1 of the Guide

Check out AI Disruptor Newsletter for more Guides on AI Tools and updates. 

How to gauge community sentiment with NotebookLM

Turn messy community discussions into clear insights with this technique.

Sometimes the most powerful discoveries happen by accident. While researching reactions to Anthropic’s latest funding news, I stumbled upon something that completely changed how I analyze community sentiment.

It’s a NotebookLM trick that most people don’t know exists, and it will improve how you gauge any community’s sentiment. (I thought about keeping this one just for me…but then I remembered how my last NotebookLM guide with Michael Spencer blew up. I want this one to do even better.)

So I spent the last few days/weeks refining this approach, and I can’t stop thinking about its implications. What used to take hours of manual analysis now takes me minutes, and the insights are significantly deeper.

Here’s what makes this particularly exciting: While everyone else is manually scrolling through endless comment threads or relying on basic sentiment analysis tools, this NotebookLM approach gives you a structured, nuanced understanding of community reactions almost instantly.

What’s different about this technique is that it doesn’t just count positive or negative reactions – it uncovers the underlying patterns, concerns, and insights that most analysis misses. And while I’ll use Reddit as our example in this guide, the principle works for understanding any community’s reaction to pretty much anything.

Before I show you how this works (video walkthrough below), let’s talk about why traditional community analysis is such a mess.

Let’s talk about the mess of community analysis

If you’ve ever tried to understand how a community really feels about a product launch, announcement, or industry trend, you know the challenge. You’re faced with hundreds or thousands of comments spread across multiple threads. Some are insightful, some are noise and trolls, and extracting meaningful patterns feels like finding needles in a digital haystack.

The typical approach? Scrolling, note-taking, and trying to piece together a coherent picture from scattered observations. Tools like sentiment analysis APIs might tell you if people are generally positive or negative, but they miss the nuanced discussions that actually matter.

What most people miss is that communities don’t just react – they analyze, debate, and often surface insights that even industry experts overlook. But capturing this collective intelligence has always been more art than science.

That’s exactly why this NotebookLM is crazy good.

It transforms this messy, manual process into something structured and insightful. Instead of drowning in comments, you get clear patterns, key debates, and incredible insights.

When NotebookLM surprised me

Here’s what most people don’t realize about NotebookLM: it has this hidden ability to understand and break down complex community discussions in ways I hadn’t seen before.

I discovered this while trying to make sense of the AI community’s reaction to a major announcement. Instead of just getting surface-level summaries, NotebookLM started revealing patterns in the discussion.

Instead of treating comments as isolated bits of text, it understands context, connects related discussions, and surfaces insights that aren’t obvious at first glance. It’s like having a brilliant research assistant who can read thousands of comments and tell you exactly what matters and why.

But it’s the ability to ask follow-up questions about the community’s reaction that I have just never experienced before with any AI tool.

Want to know what technical experts think versus newcomers? Curious about the most debated points? Just ask. NotebookLM doesn’t just summarize – it helps you explore and understand.

Breaking this down into simple pieces (How-to guide)

Let me show you how you can do this. 

6 minutes 29 seconds: 

While I’ve recorded a detailed video guide showing every step (which you should watch), here’s the core approach that makes this so powerful:

It starts with finding the right conversations.

For our example, we’ll use Reddit threads because they’re rich in community discussion. But the principle works anywhere you find detailed community conversations about topics that matter to you and have a way of extracting the text.

I wanted to analyze how the community felt about the recent news that Anthropic was raising another $2B in funding. Here is the exact Reddit thread.

Then comes the “transformation step” – it’s a simple technique that turns messy discussion threads into structured data that NotebookLM can analyze deeply.

You don’t need coding skills or technical expertise.

Just a simple trick: add .json to the end of the url.

That’s it.

Once you do that, you’ll see a huge block of structured data that looks overwhelming at first – but don’t worry about understanding it. Just copy the entire thing.

Now here’s where NotebookLM’s magic comes in.

Create a new notebook (I name mine something like “Community Sentiment Analysis -TOPIC” to keep things organized), and paste that text as your source.

Watch what happens next: NotebookLM immediately recognizes this as a structured discussion (actually exactly a Reddit thread) and starts breaking it down.

You’ll see a quick summary appear, showing you the main topics and themes it’s identified.

But here’s where it gets really interesting. Instead of just giving you a basic summary, NotebookLM lets you dive deeper through different views:

• Use the “Briefing” option for a quick overview of the community’s reaction
• Try “Deep Dive” audio overview to hear a discussion
• Use the FAQ feature to see specific questions and what people are saying

Here’s a pro tip I’ve discovered: Start with broad questions like “What’s the overall sentiment here?” then drill down into specific aspects that catch your interest. Maybe you notice there’s debate about a technical feature – ask about that specifically. Or if you see mixed reactions to a pricing change, dig into the different perspectives people are sharing.

The real power comes from building a collection of insights over time. I keep separate notebooks for different topics I’m tracking. Every time I find a relevant discussion, I add it to the appropriate notebook. Over time, you build this incredible repository of community insights that you can query and analyze.

You can ask questions, explore angles, and uncover patterns that would be nearly impossible to spot manually.

IMPORTANT: Anytime you are dealing with community discussions, do not take every statement as fact. People are wrong…a lot. This is about community sentiment, not factual analysis.

Now it’s time for you to experiment

Start simple: Pick a topic you’re curious about, find a good discussion thread, and try the .json transformation. Watch how NotebookLM breaks it down. Then start asking questions. You’ll be amazed at the insights you can uncover.

Don’t be afraid to get creative with your analysis.

• Try different types of questions.
• Combine multiple sources.
• Track discussions over time.

The more you experiment, the more powerful this tool becomes.

And here’s what I’m really excited about: I know you’ll discover uses for this technique that I haven’t even thought of. Every time I share this with someone, they come back with innovative ways to apply it. That’s why I’d love to hear about your experiments – what worked, what surprised you, and what insights you uncovered.

Remember: This isn’t just about analyzing discussions – it’s about understanding communities in ways that weren’t possible before. Whether you’re tracking product feedback, monitoring industry trends, or just trying to understand how people think about certain topics, you now have a powerful new tool in your arsenal.

Watch the video guide, and let me know what you discover. I can’t wait to see how you apply this.

Shout out to Michael Spencer and all of you in our communities. Let’s make this the most shared NotebookLM guide on Substack.

Editor’s Note

Please don’t hesitate to give a positive comment or a question, our guest and guide today is very responsive and interested in what you have to say: