I’m excited to introduce AnkiAIUtils: Explainer, a tool that provides AI-powered explanations when you struggle with Anki cards I made during medical school.
What it does:
Breaks down complex concepts into easy-to-understand explanations.
Highlights key relationships and adds helpful context.
Uses analogies and examples to fill knowledge gaps.
Why I’m proud of it:
This tool has batshit insane potential to help you understand why you got a card wrong and strengthen your knowledge. It’s like having a personal tutor for your flashcards!
Example:
For a card about febrile seizures, it explained:
A simple febrile seizure is characterized by its uniqueness and brevity during a febrile episode, which helps distinguish it from complex seizures or other neurological disorders.
How to try it:
Head over to the GitHub repo for setup instructions and examples.
Call for help:
This is a free, open-source project, and I’d love to see it grow. If you’re a developer and want to help turn this into an Anki addon, let’s collaborate!
Also, don’t forget to check out my other Anki-related repositories—I’ve got a bunch of tools that might interest you.
Slightly off-topic (as the provided example
is not convincing):
As a doctor, it still scares the hell out of me how medical students rely on simplified AI-generated explanations whose accuracy they can only assess to a limited extent. During my evaluation of generative AI, it was all there: made-up syndromes, non-existent literature, outdated information and misrepresentation. This is especially sad if you consider where we come from: From high-quality books which often have been edited by the best in their field over decades. Now it seems to be some kind of speed run through “high-yield” content for many.
I think you have a strong case against misusing the inner knowledge of large language models. And I totally agree with you. But that’s not what my projects are about! My tools are actually using the power of LLM’s on processing text, so for example, to reformulate them or make up illustrations, mnemonics, reformulate, etc. etc. But it’s really not usually heavily dependent on the inner knowledge of the LLM. This way, instead of the common pitfall of the risk of hallucination, we end up with a tool that can adapt to the optimal phrasing and presentation for every type of learner out there. Don’t you think?
