It would be really powerful if Obsidian, with appropriate permissions, applied Natural Language Processing (NLP) to my notes and offered me possible concept connections based on related ideas not just word associations. Basically, an auto complete for related ideas and concepts.
It would be great if “Obsidian AI” would be able to “explain” why it offered me these connections so I can discover unknown connections or discard the recommendation altogether. It would be even better if the connecting concepts were customized to my knowledge graph to better fit my cognitive personality. That said, I think it would also be useful if “Obsidian AI” tapped into freely available knowledge graphs to suggest other possible generic connections.
I think this feature will serve to free up our minds from the burden of making connections, or at the least obvious ones, and help us focus on higher order design and crafting of our theories, ideas, and frameworks.
Perhaps this could be a job for the Dual plugin announced here. I’m not really well-versed in it but it’s the first that comes to mind when hearing Obsidian and AI together. You could check out their documentation or join their discord
Another area to look at is using Python for NLP and other analytical tasks. While it’s early days this plugin looks interesting:
It would be good to see an open source library of scripts you could run.
I absolutely love this idea - it would really make the idea of a “second brain” more intuitive to use.
While AI-powered connections would be interesting, I’d like to have AI & NLP-powered search. In this way, one could search / ask Obsidian things like:
“What did I do last year in December” => and it shows a summary of my daily notes and important documents I created
“What are different ways of scaling Scrum?” => and Obsidian looks through my markdown notes to pull out the results that answer this