This project stems from the recognition that the modern learner's journey is fraught with distractions, inefficiencies, and manual processes. In the age of advancing AI and automation, we have the tools to make this journey smoother and more enjoyable. This project is about harnessing the power of Large Language Models (LLMs) and other software tools to streamline and potentially automate, the redundant processes involved in knowledge acquisition, consolidation and retrieval.
Nuked most of the old repo. ChatGPT workflow was fun but not very practical - UX was awful and yeah... bad. I kept the old files for now as an archive for future reference, but I have a new simpler idea. (Hopefully. Maybe. We shall see.)
TLDR
I'm looking at a much reduced scope. I want a minimum viable tool that is actually useful before trying to increase the scope.
Concretely, this means I am thinking of zeroing down on long term retention and memorisation while setting aside the rest of the overall learning components such as encoding and understanding knowledge. m thinking we keep it simple. I have a simple workflow/pipeline in mind:
- Summarise the key takeaways
- Curate the takeaways to extract element suitable for Rote Memorization.
- Generate some flashcard for each of the curated facts.
I am hoping to maybe include a proper minimal GUI this time for more control but if not possible then scripts + obsidian as a frontend also seems like an idea with potential.