A Voice-to-Text Notetaking Workflow

A Voice-to-Text Notetaking Workflow

Insight from Peter Yang

I get my best ideas when I’m going for a long walk.

The creative juices are flowing in the moment, but I always run into the same problem: 

I forget the ideas (or at least some of the important details 😅).

After many frustrating instances of this, I invested some time and research into finding a great voice -> text note taking workflow.

All you’ll need is a text -> speech app and what I call a “cleanup prompt.”

For an app, I’d recommend superwhisper. It’s lightweight, generally accurate at understanding me, and it keeps a database of all of my past recordings/notes.

I record my stream-of-consciousness thoughts on my walk, and superwhisper transcribes every voice recording into individual notes.

Now, because I’m human, those notes will be filled with “um”s and “uhh”s and “like”s. There will also be occasional misspellings, and a general lack of structure.

That’s where my cleanup prompt comes in. When I get home, I copy all of those notes and throw them into ChatGPT with the following:

Organize the notes into clear, coherent sentences and paragraphs, ensuring each idea is expressed clearly and concisely. Use bullet points or numbered lists where applicable to enhance readability and add headings or subheadings if there are distinct topics or themes. Provide a brief summary at the beginning, capturing the main ideas and insights from the notes.

You can mess around with the prompt so that it fits your preferences, but the result should be a much cleaner, organized output that you can save and refer back to 👇

No more good ideas lost to the wind 🌬️