Apparently the Whisper library is really good open source voice recognition/transcription/translation, python based. I'm imagining a tap that takes in a collection of videos or audio recordings and streams text as output. 🤯😆