<aside> 🤖 This AI transcription/summary was created on February 19, 2024. Listen to the original recording here.

</aside>

Summary

An automation system is set up to record voice memos, which are then transcribed and summarized using OpenAI's Whisper AI system and ChatGPT. The summarized text is automatically sent to Notion for further processing. The system can handle audio files up to 200 megabytes and convert rare file types to mp3. Users can receive analytics to improve based on the converted text.

Transcript

This is a test example of an automation that I set up that makes it so I can record a voice memo and put it into my Dropbox folder, which is linked inside of my Max Finder, which is essentially like all of my folders, and then it's going to automatically, the system is going to automatically detect that a new file has been added, and it is going to trigger an event in which OpenAI's Whisper AI system is going to transcribe the audio that is in this audio file that I'm recording right now, and then it's going to send that to ChatGPT to summarize what the voice memo that I'm recording right now is saying, and that summary is also going to have some key points, some questions, some sources at the end, and some things like that, and then it's going to send that whole chunk of text into Notion, and it's going to give it an automatic name based off of this, and one cool thing about this is it can be any audio file up to 200 megabytes, which is about an hour of audio, so it can handle meetings, interviews, pretty much anything, and then from there you can get analytics about how to improve, because then you just turned it into text. The only other constraints are it can be sensitive about

what you talk about, and another constraint, which I can turn that one off, I figured out a way I could turn that one off, and another constraint is there's a couple file types that it doesn't accept that are really rare, like Android phones, but those can be converted into mp3s. Anyway, this is it, this is two minutes of audio ending in two, one, two,

Additional Info

Main Points

References and Citations

Potential Action Items

Meta