Hey everyone! Models like Whisper produce significantly better transcripts compared to word-by-word voice typing. I've started using voice recognition a lot for note-taking. Here are some examples: Speaking to the mobile version of ChatGPT to copy the recognized text elsewhere, as it's much more accurate than the default speech recognition on my phone. A macOS app leveraging the Whisper model locally, allowing me to speak directly, upload audio files, or capture system audio. I use this to transcribe podcasts or videos without transcripts and to draft texts for editing later. Custom pipelines that gather all audio notes from various devices (watch, phone, computer) to create a text-based diary. I'm curious about your experiences: Do you use voice for note-taking or writing? Have you inc...
No Comments