Let’s consider machine learning first, and then automation. For each we brainstorm the potential value for note-taking.
The Value of Machine Learning for Note-Taking
- Automatic transcription of audio notes
- Speaker identification during meetings or conversations
- Resurfacing of action items, surfacing relevant information, and performing actions (e.g running a google search or sending a message) in response to audio cues during meetings or conversations
- Automatic categorization of notes
- Automatic categorization of spaced repetition flash cards
- Automatic context-filling of spaced repetition flash cards
- e.g. highlight a phrase to create a flash card; the necessary context is pulled in from surrounding paragraphs to make a coherent standalone flash card.
- Automatic connections and graph construction:
- Textual elements:
- Entity extraction from notes (“implicit nodes”, e.g. “Bank of America”)
- Attribute extraction from notes (e.g. “incredible presentation by Pierre-Antoine” gets tagged as incredible or wow (the canonical form).)
- Relationship extraction (e.g. “Diamond Age by Neal Stephenson” -> creates an “author of” relationship)
- Background knowledge elements:
- e.g. knowing that Diamond Age is a book
- Textual elements:
- Writing prompt generator / Writing assistant
- Overlap detection between people’s graphs
- The Go Note Go “clink” scenario
- Autocomplete (e.g. gmail tab completion)
- Improved spaced repetition algorithms
- Mistake detection / fact checking
- Summarization of notes
- Highlight extraction from notes
- Suggesting new ideas
- Suggesting new online resources to read / explore
- Incl. articles, papers, YouTube videos, Suggesting people to reach out to (e.g. because of their blog or Twitter or other things they’re written)
- Semantic search
- Conversational chat bots for exploring the knowledge in a notebase
The Value of Automation for Note-Taking
- Share notes with people via their preferred messaging platform
- Snooze notes and ideas so you review them later
- Set pomodoros, alarms, and reminders to stay focused and stay on task
- Note-taking via highlights (e.g. with Browserflow)
- Note-taking via screenshots (e.g. from YouTube)
- Automatic creation of spaced repetition from highlights
- Bulk refactors of note-taking
- Pulling in context from other sources (e.g. Warmly pulls LinkedIn info into Calendar, notes could pull info from wikipedia)
- Sync notes between other sources and your notebase
- Other people’s notes
- Your email client, calendar, TODO app, spaced repetition app, GitHub issues, other messaging services, etc
- Publishing of notes to a webpage (either publicly or privately), or to social media
- New data entry mechanisms: Go Note Go, terminal, email, text, browsing history, location history, terminal history, IoT devices, etc
- Spaced repetition for studying
- Inactivity reminders (e.g. “it’s been three months since you’ve done a music activity”, or “it’s been over 6 months since you’ve spoken with David Dohan”, or “it’s been over a year since you’ve written about microbiology.")