The speaker describes a personal workflow of transcribing all spoken ‘brain dumps’ and then using AI tools (ChatGPT, Perplexity) to distill and chunk those transcripts into themes, questions, and episode ideas. The transcript shows two paths for raw transcriptions: discard if not useful, or feed into AI to extract main themes, generate related questions, find similar shows and external links, and build podcast topics (example: deconstructing religion).
Chapters
Basically, I transcribe everything I say. And what's great is like, that transcribe transcription. I have two options then I can either toss it, right? It maybe doesn't. Maybe I just need to get stuff out of my brain. Or two, I can take that transcription, could throw it into something like chatgpt, Perplexity, whatever. I can just like, distill the information out of it and be like, okay, what questions did I actually ask, you know, or what you know, or what was like the main themes of what I did, right? And it'll chunk it all out for me and it'll say, hey, you want to do, like, do an episode related on. I talk a lot about, like, deconstructing real deconstructing religion in my podcast. So, like, here's the. Here's the six questions that you were kind of like asking yourself, and here's some of your thoughts on them, and here's, you know, whatever. And what I can do then is like, if you use something like perplexity or even now ChatGPT has the Internet connected to it, right? So, like, the ability to take those questions and say, okay, these are great. Help me find other questions that are in that same vein, right? And so, like, I build out kind of my questioning or topics based off of that, right? And it'll help me pull information from like, hey, find me some other shows that talk about these types of things that I can review, right? It gives me all the links so I can go back and reference them.