5. Paper Podcasts with GenFM and NotebookLM

Academic research faces a turning point. Two weeks, we discussed two AI tools: GenFM and NotebookLM. But more than the tools themselves, their impact on research behaviour caught our attention. Let me show you what I mean.

The NotebookLM shift

Google’s NotebookLM promises free document analysis for up to 50 papers. Yet, this seemingly simple tool reveals the complexities of AI-assisted research:

NotebookLM often misinterprets page numbers and citations. When Vugar tested it with legal documents, it mixed up page references while maintaining accurate content summaries. This creates a peculiar situation: correct information, wrong location. Does this actually make it easier to do your research? Since the recording, of course, Gemini Advanced 1.5 Pro with Deep Research has appeared to keep things interesting and it does a great job of finding references on the Internet.

NotebookLM’s document limits remain unclear. Unlike Claude's transparent 200,000 token limit (about 400-500 pages), NotebookLM's capacity stays a little mysterious. You upload documents without knowing if the AI processes them fully. But the podcast feature works about the same way as the following new tool. So, it’s a great way to get people interested in, for example, your research:

NotebookLM gives you different options to customize your audio overview/podcast.

GenFM changes how we digest research

ElevenLabs’ GenFM changes any PDF, so in our use case: academic papers, into 7-12 minute podcast discussions. Two highly realistic AI voices discuss your paper, creating a radio show about your research. But here’s what makes this significant:

The tool creates focused popular language summaries that capture main points without drowning in details. Each podcast maintains a consistent narrative thread asking interesting questions about the paper and answering them, something many academics struggle to achieve in their own presentations in lay language. So, it’s super helpful to get a public summary of your research paper.

Turn your research paper into a podcast in 5 easy steps.

You can download these discussions and share them alongside your papers. Think of it as a pop sci audio abstract that actually engages listeners. I think these use cases are really great for both NotebookLM and GenFM. And both tools are free.

The privacy paradox

Both tools present a critical trade-off, however. NotebookLM, as a Google product, raises questions about data use. Does clicking “delete” truly remove your research paper from their servers? Meanwhile, ElevenLabs may claim stronger privacy protections, but requires trust in a newer company (and has had a few strikes for not caring about people abusing their main product to clone voices without consent).

Consider these privacy steps:

1. Use TypingMind for local processing when possible

2. Run your own instance of Mistral for sensitive research

3. Check each tool’s privacy policy for data retention details (Anthropic is pretty good with it)

Our search behaviour evolves

AI integration changes how we search and process information. Three key shifts stand out and we discuss them in the podcast episode:

1. From keyword search to conversational queries

2. From reading to multi-modal consumption (text + audio)

3. From manual notetaking to AI-assisted synthesis

Ethical integration remains an academic challenge

Academic integrity requires thoughtful AI integration. Three principles guide effective use:

1. Verify AI outputs against source materials

2. Document AI tool use in your methodology

3. Maintain human oversight in analysis and conclusions

Some things to try after listening

1. Test NotebookLM's limits:

   • Upload three related papers

   • Request specific information that you know where to find

   • Check AI-generated references to your uploaded content

2. Create a GenFM podcast:

   • Convert your latest paper

   • Share with colleagues

   • Find out which points translate best to audio summaries

3. Audit your research process:

   • Make a list of your current AI tools

   • Document verification steps

   • Identify potential privacy vulnerabilities

The future of academic research is as much about smarter integration as it is about tools that get better every day. Focus on building systems that enhance rather than replace critical thinking. Check out our AI Research Tools webinar on the topic.