You can use it here.
🪴 Transplant pre-existing knowledge trees to your 🎋 WikiBonsai digital garden.
Have an LLM (ChatGPT) generate a semantic (sub) tree from a given word or concept. This kind of concept tree is particularly useful in a couple of scenarios:
- If you're a student, these trees can give you your bearings in relation to other, more familiar concepts. It can also point to other new, relevant concepts one needs to understand to better understand the original concept.
- If you're a teacher, these trees can save time drafting texts or curricula to someone new to the concept.
Results are viewable as an interactive mindmap and should be returned as raw text formatted in markdown outlines, which should be compliant with semtree, [[wiki-text]] among other options. This makes the resulting markdown more note-taking friendly.
The scripts are separated for cosmetic (and sanity) reasons. In essence, it is still one giant single html file so most things are happening in the global namespace. If anything ever seems confusing, sense might be made by referring back to main html file (index.html).
If this project grows further, it should probably first be refactored to use a proper front-end framework.
- Inspired by Elon Musk's quote about semantic trees.
- Neuralink semantic tree image from waitbutwhy.
- Mindmap functionality from markmap.
- Icons from icons8.
- OpenAI access referenced from tldraw's makereal template.