Robyn Anderson shows how NotebookLM cuts research time for a ChatGPT webinar. Instead of manually hunting Reddit and GitHub for prompts, she imports sources and asks the tool to generate a consolidated list. For ChatGPT Image 2.0 portrait prompts, it returned ready-to-use suggestions like "LinkedIn headshots, professional, ultra realistic, macro realism."
Chapters
I just did a few workshops on NotebookLM, and I have a really good use case for it today that I wanted to share because people really like NotebookLM, but they're not like very clear on how or when they should use it. So today I'm hosting a webinar on ChatGPT 2.0. Now I could spend the rest of my day searching Reddit, GitHub, the internet, LLMs for ChatGPT 2.0 prompts Or I could use NotebookLM to find and import these sources for me. And then right here in the middle, I can go and ask questions. And then I could also go and just create a document. So I just had it create a whole overview of prompts. So here in the middle, I can chat against all these sources and say, what are good ChatGPT Image 2.0 prompts for portraits? And it'll just pop right up. So here it's giving me like LinkedIn headshots, professional, ultra realistic, macro realism, which these are all really great. And once you see them, you want the prompt. So basically you should use NotebookLM when you're trying to learn something new and there's tons of information out there about the thing, but you don't have hours and hours to read or research the thing because honestly, who has time for any of that?