I’ve been working on an open-source document format / viewer idea I’m calling Adaptive Markdown.
The basic idea is: instead of a document being static text it's controlled by coding agents.
You interact with the document more like a live workspace. This has different implications depending on what you are doing.
I made a short video demo here:
The thing I’m most excited about is academic / technical reading. In a few years I don’t think people will just read papers passively. I think they’ll translate passages, ask questions, generate examples, explore alternate proofs, run code, attach notes, convert math to Lean when possible, and keep all of that inside the document instead of scattered across chats and notebooks. This is trivial to do inside a browser with coding agent that has access to JS, CSS etc.
Some possible use cases I’m thinking about:
-Turning articles and books into personalized learning objects
- lecture notes with automatically maintained structure
-documents with embedded code, tables, consoles, images, audio, or video
-AI-generated alt text and descriptions
Incorporate Adaptive Markdown into automated work flows
eventually, things like automatically recording audio in lectures and taking a picture of a blackboard and turning it into LaTeX notes inside the document
It’s very early, but the workflow already feels surprisingly useful to me.
GitHub: https://github.com/SemiSimpleMath/Adaptive-Markdown
Curious whether this seems useful to anyone else, or whether I’m just overexcited because I built it.
So far it's only configured for Anthropic coding-agent SDK, but in couple of days we will have it running on Codex as well.
[link] [comments]