When I started building my digital garden in Obsidian, I wanted more than just a place to collect notes. I wanted a space where ideas could grow, connect, and evolve. Over time, this approach has shaped one of the most dynamic parts of my garden: the AI Index.
The AI Index is my living glossary of key concepts in artificial intelligence and machine learning. It’s where I synthesize what I’m learning through Coursera courses, podcasts, blog posts, lectures, and my own teaching. But what makes it truly powerful, and hopefully useful to others, is how it leverages interactivity through Obsidian features like backlinks, graph views, and Dataview queries.
These tools have transformed my notes from static entries into an exploratory, evolving resource. Here’s how.
Backlinks: Connecting Ideas Across Contexts
One of the first features that made my garden feel alive was Obsidian’s backlinks. As I study AI topics, like neural networks, bias in AI, or transfer learning, those terms naturally show up in multiple places. This could be lecture notes, article reflections, workshop plans, and blog drafts.
Take the Generative AI page as an example. In my Obsidian vault, I opened the note and saw two linked mentions in the backlink panel. The first one pointed to the AI Index, which made sense, since it’s one of the core entries. But the second backlink surprised me: it connected to a note I took years ago on the socio-ethical challenges of Generative AI. I had completely forgotten about it.
Backlinks in Obsidian reveal connections across time and context. Here, the “Generative AI” note links to both the AI Index and a long-forgotten reflection on socio-ethical challenges, surfacing ideas I didn’t even remember I had.
That’s the power of backlinks in Obsidian. They surface connections I didn’t plan for, helping me revisit and deepen ideas over time. By linking terms like Generative AI to index entries, blog posts, course reflections, and workshop planning documents, I’m building a web of associations that evolves as I learn.
And this isn’t just helpful for me.
In the published version of my digital garden, these backlinks appear as Pages mentioning this page. For readers, this turns a single note into a springboard. They can follow their curiosity, click through connected notes, and uncover how these concepts interact across different contexts. It’s a way to move laterally through knowledge, just like I do.
In the published garden, backlinks appear as “Pages mentioning this page.” These connections invite readers to explore laterally, following threads of meaning and making their own discoveries.
Visual Graphs: Seeing Structure Emerge
Another way I support discovery is by using local graph views. In the AI Index, major concepts like machine learning or ethics in AI serve as hubs, connecting to more specific terms like supervised learning, loss functions, or explainability.
Graph view is one of Obsidian’s most eye-catching features, especially when you’re just getting started. While it’s often treated as a novelty, I’ve found it most useful as a way to visualize the interconnectedness of my thinking.
In the example below, you’ll see the full graph view of my vault. Each dot represents a note. Notes that are closely related, through backlinks or shared topics, naturally cluster together.
You can spot clusters for blog posts, newsletter issues, books, podcasts, and thematic topics. I’ve highlighted the cluster for my AI Index, where you can see a dense web of direct connections between entries. But what’s just as interesting are the unexpected links, connections to the index that were made elsewhere in my notes. These stray connections often lead me to new insights or forgotten ideas, an invitation to go explore.
Each dot is a note, and each cluster a conversation. The AI Index cluster reveals both intentional links and unexpected discoveries.
These local graphs help surface the structure of the index without overwhelming the reader. It’s especially helpful when I’m presenting complex material: instead of reading a long definition, someone can see how a concept fits into a broader conversation.
This visual layer supports intuitive navigation, something that’s often missing from traditional glossaries or course notes.
Dataview Queries: Surfacing What Matters
Perhaps my favorite tool in the garden is Dataview. Dataview looks at your notes, not just the names, but the information inside your notes, and finds the ones that match your request. Then it builds a list, table, or calendar for you automatically!
Imagine each note in my Obsidian vault as a trading card. I’ve got cards about learning, cooking, music, books, and podcasts. All covering a wide range of topics. Some of these cards are about AI, machine learning, cognition, assessment, and evaluation. Within the AI group, I might have one for Generative AI, another for Ethics in AI, and one for Bias in AI. These notes live in different folders like “Podcasts,” “Webinars,” and “Books,” depending on where the idea came from.
This is where Dataview becomes powerful. It acts like a super-smart card sorter.
I can ask it to gather all notes related to AI, no matter where they’re stored. It builds a table or list that lets me quickly scan, review, and identify connections between ideas. If I notice a gap or an outdated note, I can jump in and revise. This not only helps me make better connections for myself, but it also improves the experience for anyone exploring my published digital garden.
For example, I use Dataview to:
- Highlight entries marked 🔹 beginner-friendly or 🔸 advanced
- Automatically surface recently added or updated concepts
- Generate topical clusters (like everything related to Training & Optimization)
- Pull together curated resources for workshops or upcoming teaching units
The image below shows a Dataview query from my Obsidian vault, listing the last ten notes tagged with AI, machine-learning, or generative-ai. Each entry displays the note’s status—whether it’s a seed (a new idea), a plant (developing), or an evergreen (fully fleshed out). I can quickly click any link to jump into the note and explore how it connects to whatever I’m currently working on. For example, I might decide that a note on what “good” AI looks like would be great for discussion in my upcoming workshop. But something like the recent mapping of a fruit fly’s connectome (a brain-wide, synapse-level map) might be too niche unless we’re diving deep into how neuroscience intersects with machine learning.
When I’m preparing for my summer AI workshop, I can tell Dataview:
“Hey Dataview, show me all notes tagged CSPD-Workshop, including entries from Coursera, my reflections, and my AI Index.”
It then builds a dynamic table that updates itself whenever I add something new. Whether I need a list of key terms, a reading log, or a set of questions, I’m still thinking through. Dataview keeps it organized and interactive.
It’s like having a digital research assistant that helps me learn, reflect, and share more effectively.
So instead of hunting through all your notes, Dataview builds a custom dashboard just for you and updates it every time you add something new. As my knowledge vault grows, Dataview helps me organize and present content based on tags, sections, and custom properties. This turns my garden into a live dashboard for learning, planning, and teaching.
Designing for Discovery
What makes this garden “interactive” isn’t just the tech, it’s the intent behind it. I’m designing with discovery in mind. Readers can follow threads of curiosity, see how ideas connect, and watch concepts evolve over time. It’s more like wandering through a well-labeled forest than flipping through a textbook.
I don’t expect someone to read the AI Index from top to bottom. Instead, they might start with a blog post on AI in education, jump to the glossary entry on bias, follow that to fairness in AI, and then browse related teaching notes. That kind of serendipitous exploration is what makes this garden feel alive, for me and for others.
Want to Explore?
Here’s a peek at the AI Index, which includes:
- Clear, accessible definitions of AI/ML terms
- “Explain Like I’m Five” (ELI5) breakdowns
- Contextual links to course notes, workshop planning, and external resources
- Dynamic sections like Core Concepts, Training & Optimization, and Ethics & Challenges
🔹 = Essential / beginner-friendly
🔸 = Advanced / nice-to-know
Check it out here, and feel free to browse, click, and wander. This is a living document, and I’ll keep adding to it as my understanding of AI grows.
Final Thought: Publish to Learn
Publishing this work-in-progress has changed how I learn. Writing for an audience, even a hypothetical one, pushes me to clarify my thinking, document more intentionally, and build with care. Interactivity isn’t just for the reader. It’s for the writer, too.
The use and development of a digital garden allows me to write and share, but also to show my notes and leave these connections public for others to check out. As I teach a class, publish a paper or blog post, or talk with someone at a keynote or webinar, I can quickly connect them with my notes to answer questions and encourage their exploration.
If you’re building a digital garden of your own, I’d love to hear how you’re using features like backlinks and Dataview to support deeper learning and open sharing. Let’s keep building knowledge together, and in public.