Filing this one under the "contributing to the community" category: I pulled together a YouTube video to illustrate a neat way to go from audio recordings on any standard audio recording device, to a best-in-class transcription -- run through Whisper (openAI’s text-to-speech) and then GPT to further clean up the results -- for free, all in a single command: "-transcribe" The 3 prerequisite steps to doing this are (1) installing AutoHotKey; (2) installing Whisper on your computer; (3) running the AutoHotKey script. [ All necessary resources are available in video notes ] Once you have done this, then you simply record your audio file, plug it into your computer, and then type that command above and voila the rest is done for you. A lot of different directions one could take the underlying building blocks, but a fun little tool that I thought could be helpful for other people using tools like this. Thoughts/suggestions/comments welcome... those who know me know I spend a lot of time tinkering with these tools. This is a tiny reflection of that! And sorry for not sharing my screen as early as I should have!
How to convert audio to highly accurate text for free using AutoHotKey, Whisper, and GPT
https://www.youtube.com/
Nikhil Karnik this is interesting 💡
Superhuman Dan!
Conduct research to increase consumers' financial well-being
1moWould this be appropriate even for highly sensitive recordings?