23 Comments
User's avatar
Sam Illingworth's avatar

Thanks team, this is an excellent post. I've just moved over to Gemini from ChatGPT, so I'm going to be using this myself as I think about both the direction of slow AI and also my own AI research into the new year. And Santa would definitely place both of you on the good list. Well done. 👏

Raghav Mehra's avatar

Thank you Sam for your constant support! Yes, I too, am making the switch from ChatGPT. A toolkit of Claude, Gemini and NotebookLM, and Perplexity is more than sufficient for most work. Happy to see what's coming in Slow AI in 2026 (especially our first article collab😉). Have a great festive period!

Sam Illingworth's avatar

Cheers man. You too. I’m sure there'll be loads more content before then though as well. 🙏

Hina Gondal's avatar

Beautifully written ❤️

Raghav Mehra's avatar

Thank you Hina! :)

Manisha's avatar

Great post. I have been using NotebookLM a lot for learning too, never thought of merging it with Gemini. Something I have to try, thanks!

Raghav Mehra's avatar

Thank you! Yes, merging your research documents in Gemini and using Canvas/Nano Banana there to generate visual assets is an excellent value-add. Let me know how your experience is!

Modern Mom Playbook's avatar

I sent this to my husband! He is a PM at Google!

Raghav Mehra's avatar

Wow, thats amazing, thanks Anastasia. Curious for his take on this!

Daria Cupareanu's avatar

So much great stuff in here!

Raghav Mehra's avatar

Thank you Daria, appreciate your feedback!

Darlene's avatar

This was a wild ride, from Santa to enterprise tech to bananas, but I appreciate the central point: AI is best when you use it as a system, not a one-off trick.

I’m curious: do you think NotebookLM is worth learning if someone’s already comfortable with other tools?

Raghav Mehra's avatar

Hi Darlene! Thank you and you are right—it's more about a system that suits your work requirement that an stand alone tool. I think NotebookLM is great tool for documenting your research in an organized way and having insights broken down to you very seamlessly. Having said that, I'd love to know what tools you use for the same workflow!

The AI Architect's avatar

Incredible walkthrogh. The wealth management use case is genius because it highlights what people actually struggle with: not finding info but synthsizing it under constraints. Showing how NotebookLM's cited synthesis feeds into Gemini's contextual recommendations is practicly the most useful thing I've seen for decision frameworks. One question though, when you're building these across multiple sessions do the notebooks stay persistent or do you rebuild context each time?

Raghav Mehra's avatar

Thank you so much! Yes, the wealthtech use case was very appealing given my professional domain is related to it. NotebookLM and Gemini do a great job together and separately depending on what your workflow demand is. If you are within the same notebook, you can maintain the context by loggin in it everytime. If you want to tinker the subject a bit and want to use Studio for the revised context, I'd recommend weeding out the links that are no longer needed.

Karo (Product with Attitude)'s avatar

Thank you both! I'm also using Gemini (and Claude) more and more, so this is very useful. Also, I liked ''What Santa Would Say'' 😂

Raghav Mehra's avatar

Thank you Karo! Means a lot you appreciating our work. Yes, these tools are more than useful and I use them extensively now as well. Haha I wanted to make this article Christmassy and spend a lot of time in making consistent parallel Christmas narratives! 🎄

Ileana's avatar

So useful and fun! 👌

Raghav Mehra's avatar

Thank you Ileana, glad you liked it! 🤗

Mia Kiraki 🎭's avatar

The reindeers approve of these workflows! Love the twist and they're super applicable to us, day to day humans hahah 😂 thank you guys! ❤️🎅

Raghav Mehra's avatar

Haha thank you Mia! I'm sure Santa has you on a 'super-nice list' and reindeers will deliver you gifts that you deserve! 🎁🦌

Mia Kiraki 🎭's avatar

He better hear you!!

Raghav Mehra's avatar

Yes, its always about a system than a tool in itself. About NotebookLM, I do think it has a niche use case for handling multiple sources of research and documentation suites. Like most AI tools it's not irreplaceable but I haven't found any another subsititute for its unique capabilities.