Build AI-Powered Research Systems with NotebookLM and Perplexity

[@BlazingZebra]

🤗Grab the Cheat Sheet in Patreon (all prompts, resources etc) - www.patreon.com/posts/notebooklm-for-106794985

[@Custodian123]

1:25 important to note, Warren is focused on really understanding what he is looking at. Less about staying ahead of the market.

[@BlazingZebra]

true, good catch

[@JeffRobinson65]

AMAZING TUTORIAL ON NOTEBOOKLM!!! I've been using it for a while, but you've given me even more tips to make the most of this tool

[@BlazingZebra]

killer! glad u found it helpful!

[@AndreAloisioToledo]

A great tip is to use Perplexity as a source for articles and other information, and then ask specific questions and cross-reference data in NotebookLM.

[@BlazingZebra]

exactly! perplexity for finding sources, notebooklm for insights and analysis.

[@plarpco]

Thanks for this one! I hadn't tried notebookLM before, and its turning out to be very useful. Cheers!

[@BlazingZebra]

Glad you found it useful! XO!

[@azRome1990]

Thank you for the video! Can you upload books (pdf or pdf ocr) on it (thousand pages)?

[@BlazingZebra]

pdf is the way to go, each source maxes out at 500,000 words. a thousand pages might be pushing it, u can break it into a couple sections tho.

[@abderrazakcroxup321]

thank you very much for all your efforts

[@BlazingZebra]

@@abderrazakcroxup321 hey! thanks for watching!

[@billgrey]

In your estimation, is this similar to Claude's new "Projects" feature. Have you looked at that yet?

[@BlazingZebra]

I would say it is similar but might have slightly different use cases. Claude’s projects feel to me a bit more like openai’s Custom GPTs (you can add files to those as well). NotebookLM seems like it excels at research, while the others might be better for specific workflows. I am messing with Claudes new features more this week. Stay Tuned ;)

[@nighshadeai]

But isn't this the same as making a GPT with the same info?

[@BlazingZebra]

this is a common question i keep getting. there are a few major differences. first the user experiences geared in a much more friendly way for finding insights and collecting notes. Secondly, the LM is grounded in the source material. That’s not the case with the custom GPT’s in that can be a major problem. Thirdly, I believe the rag system used to access. The sources is much better than custom GPT. It also supplies citations to exactly where in the source material it’s gathering its information.

[@heesongkoh]

The amount you can upload count.