In spaced repetition, it's tempting to ponder per-card cost—"should I pay three minutes over my lifetime to memorize this?" But for me, marginal cards are effectively *free* to add: 1. I review cards in time that'd be dead anyway. 2. I can't add enough cards to fill that time.
-
Show this thread
-
For instance, often I'll review cards while waiting 10 minutes for the next train. This doesn't "cost" me 10m because I would've just been screwing around on Twitter or something. And even with 1000s of cards, it often takes less than 10m! I can't write new ones quickly enough!
1 reply 0 retweets 4 likesShow this thread -
This leads to a *very* unintuitive feeling: it's… effectively free to memorize as many new cards as I can write? What? Which in turn means that for me, the real marginal cost is in the moment of *writing* a new card.
2 replies 0 retweets 9 likesShow this thread -
I think there are ways to mitigate that cost, but some of it's essential: in writing cards, I'm synthesizing, comparing, filtering, personalizing, etc. I'm interested in structures and routines which support me in doing those things more rapidly or with less friction.
2 replies 0 retweets 6 likesShow this thread -
Once one's attention is on the writing-a-card cost, environment becomes a key barrier. The cost is way higher on a smartphone keyboard, so I make cards on my laptop. But my habits often misalign: I read while out, I read in bed, etc—the laptop's not around. Gotta "buffer" cards…
3 replies 0 retweets 5 likesShow this thread -
Replying to @andy_matuschak
I’ve been struggling with the card-making bottleneck lately too. Tried entering them in a spreadsheet yesterday + using import. Seemed to work better than the app, but it’s no panacea.
1 reply 0 retweets 1 like -
Replying to @KevinSimler @andy_matuschak
I’ve also considered writing an email-to-anki script: subject becomes question, body becomes answer. This would allow me to use any email interface to create cards, e.g., Siri. But I despair of all the plumbing that would be required to make it seamless.
1 reply 0 retweets 1 like
Last thought is to use textbook glossaries to create vocab cards in bulk. I realize this violates @michael_nielsen’s advice to make one’s own cards, but I imagine it a bit like pretraining a neural network: just learn some stock info of middling relevance, then refine later.
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.