AI Assisted Search / Q&A / Chatbot / Text Generation

Feature requests
21

I never had the opportunity to play with GPT-3, but I’m very curious how this could fit the Note-taking picture.

I’m pretty sure that many people would feel great when being helped to generate specific types of repetitive content, like some office work, forms, things with known structure.

Perhaps using it in a personal vault would generate lots of noise and ‘distant’ text. The note-taking activity is rather intimate, needs emotional involvement; here, more is not better (at least for me).

But it would be cool to give it a try and discover when one could use this.

I would like something that helps me diminish complexity, which proposes to delete some notes, merge two notes that are too similar, summarize, etc.

22

I’m curious with what are you doing with your bookmarks?

With obsidian notes I’m experimenting with showing text of other notes that is similar to my current text. Also doing page-rank to my notes. It’s really useful.

23

I’m building a recommendation engine for my bookmarks. Right now it’s based on TF-IDF. I tried fancier embeddings like bert and doc2vec, but I’ve had the best results with TF-IDF.

With this similarity metric, I’m trying things like building a graph based on the top 5 most similar and then running some type of clustering on top of it.

It’s actually working surprisingly well and was born out of me not wanting to carefully organize every blog post, article, and paper I find interesting. It’s gonna be the first filter I use to determine if I want to distill an information source into my obsidian knowledge base.

It’s cool to hear your also using ML to understand your notes I think there’s a huge opportunity for this!

24

like, existing notes that may relate to what u are currently editing

25

Yes, a similarity metric for note content and a merge tool could immediately help to find and adjust notes that may be redundant.

26

For example, the ‘Sugested Topics’ section in this forum already is very effective in showing related content. How it works?

2 Likes
27

Text Generator Plugin’s goal is to have at the same time the capabilities of Obsidian with all my notes and the capabilities of Open AI GPT-3, which hopefully makes my writing very fluid, fast, and powerful.

28

I don’t know a lot about how natural language processing works and how models are trained but I wanted to share anyway because I thought this could be a game changer for obsidian especially as a tool marketed as a second brain. It would be great if a natural language model can be trained using one’s knowledge base from the md files stored in their vault and then an AI can respond to any questions the user asks based on that knowledge.

I think this would be the ultimate and most effective method for information retrieval and it could also allow users to identify creative links between their ideas that might otherwise not have come to notice.

3 Likes
29

I consolidated a few FRs/plugin ideas here. We are aware that there is hype/demand for this and, believe it or not, this is not technically challenging provided that you are willing to send your data to ClosedAI. In fact, I think we are getting closer to ten plugins that interface with OpenAI.

If you want to do things locally (Obsidian’s ethos) and cross platform on a regular computer or smartphone, things get much tricky. We’ll see where this goes.

5 Likes
30

I built sth similar recently. You may find it useful for your project GitHub - rpast/obsidian_vault_neighbors: Semantic connections recommendation script for Obsidian.md notes.

2 Likes
31

For anybody wanting to kick start their own script for conversational interface over given notes with OpenAI API, here is a notebook I’ve just banged out to see if their cookbook is worth anything. It turns out it is.
The script takes just one pdf file, applies bespoke transformations to extract chapters, calculates the one having smallest distance with the user’s query and sends it as a context to gpt3.5 API call.
Just a poc, but may be useful for anybody wanting to build their own stuff. GitHub - rpast/tsunset

2 Likes
32

is a nice demo of NotebookLM, including Readwise integration (via its Google Docs export)

1 Like
33

is a similar open-source product,
http://memora.app
I think a careful plugin in Obsidian could get pretty close to this…

does have an Obsidian plugin but it looks like just a dumb frontend to their API

1 Like
34

eagerly waiting!

2 Likes
35

Hello there. I would like to know if Text Generation works with other languages instead of English. Thanks.

36

I can verify, Khoj is pretty incredible! It’s incredible to be able to chat with your daily notes and find stuff incredibly easily with the A.I. search.

37

This is now Obsidian’s achilles heel.
Local AI needs to be a core plugin.
Something like smollm2 is very light.
The functionality could be READ-ONLY for safety.

No assumptions about AI privacy: Read-only and offline only.

Until then, you can point

at your vault and ask questions there. It’s a bit easier to setup than Khoj.

I hope this helps :slight_smile:

2 Likes
38

I have been a long-time user of Obsidian, but this is my first time registering on the forum because I am both excited and apprehensive about the current state of Obsidian. Even before the AI era, its powerful openness and intelligent interaction logic made me deeply fall in love with it. As a researcher, my primary need is to record plain text notes, sometimes inserting handwritten content or images, and Obsidian meets these needs perfectly. Even when I need to write text in formats other than Markdown, the plugin marketplace offers various powerful text-editing enhancements that provide an experience with a lower learning curve but no less—or even superior to—Vim. For example, one of my common use cases is writing LaTeX in Obsidian and compiling it directly with a LaTeX compiler that supports .md files.

What pleasantly surprised me is how seamlessly Obsidian has integrated with AI in this new era. If before the AI era I would modestly say Obsidian was one of the best personal note-taking apps in the world, today I can confidently say it is the best. Its Markdown content serves as the best retrieval source for AI, its canvas feature feels like a workflow stage tailor-made for AI years in advance, and its bidirectional linking natively enhances AI’s understanding of the knowledge base like a knowledge graph. All these foundational advantages of the AI era are quickly realized as usable features thanks to its open and vibrant plugin ecosystem. I have already joyfully downloaded several AI plugins that significantly improve my note-taking experience and can’t put them down. This is precisely what excites me and one of the reasons I registered for this forum to write this post.

However, another reason driving me to register is my unease. Obsidian is undoubtedly the best among pre-AI-era note-taking apps in embracing AI, almost solidifying its position as the best personal note-taking software today. But its AI ecosystem is not perfect—there are still current issues and future concerns. The most immediate problem is that, despite Obsidian having the best foundational advantages for AI, the official team hasn’t packaged these features for us directly. Everything still requires plugin developers and users to implement them manually. For example, with AI plugins featuring RAG capabilities, each one I install requires rebuilding the database using embedding models, which undoubtedly leads to significant resource redundancy and waste.

Additionally, different third-party plugins offer varying levels of support for various model providers. Users with only a few provider APIs may struggle to ensure the next plugin they download will work properly. Even for those with access to multiple providers, manually configuring each plugin’s settings adds substantial overhead to the user experience. I believe the official team should address these issues with better services:

  1. Provide unified database access support: Since different AI plugins have varying database requirements, the official solution shouldn’t be a single database service but rather a “database hub” where users can create a single database for all plugins to share or multiple dedicated databases for specific plugins.
  2. Offer unified LLM API management: The specifics of this feature don’t need much elaboration, as there are already many examples in the market.

These are the immediate issues I think need addressing. As for the longer-term outlook, I am optimistic, but I also see potentially fatal challenges that may be hard to resolve. Before the AI era, Obsidian thrived thanks to its open plugin ecosystem, becoming one of the best note-taking apps and quickly adapting to the AI era. Though I’m not a professional developer, I suspect that maintaining compatibility with existing third-party plugins could become a burden for Obsidian’s next leap forward. Currently, all its advantages are relative to pre-AI-era note-taking apps, but what might dethrone it won’t necessarily be its past competitors—it could be future challengers. Newer apps, natively designed for AI from the ground up, could absorb all of Obsidian’s strengths while being lighter, more efficient, and unencumbered by legacy support. However, because Obsidian must maintain compatibility with its plugin ecosystem, making major architectural changes won’t be easy—like a large ship struggling to turn. This could prevent it from making crucial adjustments, leading to failure in the distant future. While I recognize this problem, I must admit I don’t have concrete suggestions, as I’m just a user more familiar with the experience side rather than technical implementation. Still, I hope the Obsidian team takes this seriously and begins strategic planning early. I want to continue happily editing my notes, papers, and code in Obsidian years from now without needing to migrate to another platform.

Since this is my first time posting here, I’m unsure if my thoughts overlap with others’. If so, moderators are welcome to delete this post.

English is not my first language, so I appreciate everyone who took the time to read this.

Here’s to hoping Obsidian delivers an even better experience in the future!

3 Likes
39

I’m among those who’d immediately start searching for a new notes app if AI every made it to core Obsidian.

2 Likes
40

Same here. No thanks.