Give us any document — a chapter, a paper, your class notes — and get back a guide that quizzes you while you read, remembers what you missed, and brings it back right before you'd forget.
Quick one: what builds stronger memories — re-reading your notes, or closing them and testing yourself?
Testing yourself — by a wide margin. Psychologists call it the testing effect, and it's one of the most replicated findings in learning research.
That flip you just did? That's the whole product.
The problem
Reading alone doesn't stick
You've probably had this happen: you read something genuinely good, felt like you understood it, and a month later could recall roughly three sentences. That's not a personal failing. In Why books don't work, Andy Matuschak argues it's the default outcome — prose quietly assumes that people absorb knowledge just by reading sentences, a theory of learning that cognitive science discarded long ago. The readers it does work for are doing invisible, effortful extra work the medium never helps with.
So instead of blaming readers, change the medium. Matuschak and Michael Nielsen's Quantum Country did exactly that: a textbook where quick memory checks are woven into the text every few minutes of reading, then return on an expanding schedule over days and weeks. Their readers held onto detailed technical material with strikingly flat forgetting curves. This tool applies that design — plus the wider evidence on retrieval practice, spacing, and active learning — to any document you give it.
The evidence
Where these ideas come from
Matuschak (2019) — Why books don't workThe diagnosis: books and lectures share a faulty implicit model of learning; media can instead be designed around how people actually learn.andymatuschak.org/books →
Matuschak & Nielsen (2019) — Quantum CountryThe prototype "mnemonic medium": retrieval prompts inside the text plus a spaced schedule, with data showing remarkably durable retention.quantum.country →
Matuschak — How to write good promptsThe craft this tool encodes: each prompt should test one thing, precisely, with real retrieval effort. Big multi-part questions let you feel fluent while remembering little.andymatuschak.org/prompts →
Roediger & Karpicke (2006) — Test-enhanced learningThe classic demonstration that practicing recall beats re-studying for long-term retention — even when the re-studiers feel more confident.doi:10.1111/j.1467-9280.2006.01693.x →
Cepeda et al. (2008) — Spacing effectsLarge-scale study of how review timing shapes retention; the empirical basis for expanding-interval schedules.doi:10.1111/j.1467-9280.2008.02209.x →
Dunlosky et al. (2013) — Improving students' learningA review of ten study techniques: practice testing and spaced practice rank highest; re-reading and highlighting rank near the bottom. (Yes, we see the irony of our highlighter theme.)doi:10.1177/1529100612453266 →
Freeman et al. (2014, PNAS) — Active learning meta-analysisAcross 225 studies, actively engaging beat listening by about half a standard deviation and cut failure rates by a third — the case for the do-it and teach-it exercises.doi:10.1073/pnas.1319030111 →
What you'll get
A guide that quizzes, explains, and waits for you
Your document comes back reorganized into short, readable sections. After every few paragraphs, a small stack of flashcard-style questions checks what you just read — answering them is your first repetition, and every answer joins a spaced-repetition schedule saved privately in your browser. Each section ends with one genuinely hard multiple-choice question (with the reasoning explained) and a teach-it exercise where you put the core idea in your own words. Key terms are highlighted and tappable for plain-language explanations. The whole thing downloads as a single file that works offline, forever, for free. See a live sample →
Try it
Make your guidetakes about a minute · nothing is stored on our servers
Drop a file here, or click to choose
.pdf · .docx · .txt · .md — or paste below. Works best with explanatory writing, 1–15 pages.