Using Notebook LM for the first time

I’ve ignored Google’s AI products in favor of ChatGPT and Claude.

But watching Tiago Forte’s video on the topic made me realizing that I was sleeping on something very powerful.

So I gave it a test run.

I uploaded 11 sources relating to Reinforcement Learning and the role it plays in DeepSeek’s algorithm.

It then synthesized all the sources into a podcast between two AI hosts that followed the specific custom instruction:

I want to learn about reinforcement learning so that I can better understand the innovation behind DeepSeek. I am not technical and don’t know statistics or computer science, but I am a “smart layperson.”

Here’s the resulting podcast. LMK in the comments if it communicates the idea well.