Flashcards are "dead fish": rethinking spaced repetition around highlights

The essay argues that flashcards are a limiting foundation for memory systems. They work well for discrete facts such as vocabulary or country capitals, but become harder to create and less effective as learning moves toward complex ideas. A card’s wording can become such a strong cue that a learner recalls the expected answer without being able to use the underlying idea flexibly. The author therefore distinguishes the goal - reliable, useful memory - from the particular technique of spaced repetition, and calls for systems that draw on a wider understanding of how people remember and learn.

The proposed direction is to make highlights, rather than user-authored flashcards, a central part of the system. A highlight marks material a reader finds significant and keeps an idea connected to the passage and its surrounding context. Machine-generated prompts could use that material to create opportunities for review without requiring readers to translate every insight into a standalone question-and-answer card. The broader ambition is an enabling environment that helps readers retain and apply what they read, not simply rehearse isolated facts. This is presented as early design work toward a different kind of memory system, rather than evidence that a finished approach has already been proven.