Flashcards, menmonics, and spaced repetition systems are the most efficient way for students, researchers, and the curious to develop their knowledge of new subjects. Despite this being well established in research and the help of computerised systems like anki, there are several user experience problems which lead to people underutilising the technique:
- Time consuming to produce: Even if they know that flashcards are the best system for learning a topic, often users don’t use them because they are so time consuming to produce (even though they may take very little time afterwards)
- Bias towards simplest to produce content: Flashcards are underutilised for learning topics outside of language learning, because vocabulary drills are the easiest kind of flashcards to make. By using AI it's possible to expand the use of mnemonics such as rhymes and acrostics to aid learning in a way that previously was very time consuming.
- Stumbling on old content: When practicing cards over a long period, forgetting the answer to an old prompt can be challenging because the answer on it’s own doesn’t provide enough context or reasoning to remind the user why the answer is what it is. e.g. you forget what “je m’appelle” means in English, but when you read that the answer is ‘My name is’, this doesn’t explain the grammar rules of why this is the case. This is a big problem because it can make using flashcards for long term recall redundant.
AI Flashcards extension solves these UX problems in multiple ways:
- Use a large language model to generate flashcards quickly and automatically from any text input.
- Leverage LLM capabilities to generate a wider variety of prompts and mnemonic devices than was previously feasible in a spaced repetition system. e.g. multiple choice questions, acrostic poems, rhymes, etc.
- Track the source material of all flashcards, so that it can be easily and immediately revised whenever something is forgotten.