AI Makes Note-Taking Obsolete

The real shift is from better capture tools to software that does the remembering and synthesis itself. In Jeff Tang's framing, notes are only a rough input into thinking, not the end product. Once models can ingest calls, docs, and web pages, find the common thread, and return the answer in plain language, the user job moves from writing things down to reviewing, steering, and acting on machine generated context.

• Athens was built around graph based note organization, with links, properties, and relationships helping users structure messy ideas. Tang's point is that AI changes the layer above that, because summarization and retrieval from unstructured text can bypass the manual step of turning thoughts into neatly organized notes.
  1 sacra
• This is the same leapfrog pattern showing up across productivity software. Zapier's AI push was not about making traditional workflows slightly easier, but about replacing trigger by trigger setup with a natural language assistant that can act across apps. The interface shifts from configuring the system to asking the system.
  2 sacra
• The winners in this kind of market are less likely to be the apps with the prettiest note editor, and more likely to be the products closest to the raw source material, meetings, documents, inboxes, and chats. That is why Otter faces pressure from Zoom, Teams, and Google Meet, which can build notes and summaries into the meeting itself.
  3 sacra

Going forward, productivity software keeps moving upstream, from helping create artifacts to removing the need for those artifacts in the first place. The durable products will own ingestion, retrieval, and action loops around real work, not just the screen where a user types notes after the fact.