I recently went down a beautiful rabbit hole reading the SuperMemo wiki. The creator Piotr Wozniak has been thinking deeply for decades about optimizing learning with the help of software. His theory of memory seems so self-evident and intuitive, and yet somewhat uncommon. I had heard of SuperMemo the software as the OG spaced-repetition app, but that it was a bit niche and that Anki was preferable, so I never looked more into it until now.

The pinacle of SuperMemo isn't actually spaced repetition flash cards, but a workflow called incremental reading. It entails collecting your reading into SuperMemo, maintaining a priority list and then reading what the system presents to you according to its optimization of spacing and engagement. You read a few pages of an article or book, extracting the snippets that are most important, and moving on to different reading material (reading in increments). The snippets get scheduled for later review, and after you see them a couple times, you transform them into cloze deletions¹. You therefore digest your reading into spaced-repetition.

Incremental Reading combines a reader app in which you collect all sorts of articles and books, and an SRS. It also accomodates a reading style that involves "interleaving" many different reading materials. This easy jumping (they say) maximizes your attension, engagement, and benefits your memory. When something starts to feel like a drag, just move on to the next one. By picking up where you left off in a book you do a kind of subconcious review of the last section– this is already begining the "repetition". Because you've been extracting things as you go, you can also review backwards if you need to, although this screws up the timing of your reviews.²

At the very least we can see that this workflow encourages/requires that you actively think about what information you really want to keep. The system also seems like it makes the reading/review process comparatively engaging to the usual scroll-addict feeds; by keeping track of your reading history and priorities, by encouraging you to just skip past what doesn't interest you or reschedule it for later, and by scheduling your reading increments based off of your engagement³. Instead of accomplishing that with slot-machine dopaminergic tricks, its from the organic reward of accumulating durable knowledge.

This is programmable attension, this is a practice to replace scrolling. I'm very excited about building my own version.