Definition:

Piotr Wozniak (born 1963) is a Polish researcher and software developer, best known as the creator of SuperMemo and the SM-2 algorithm — the first computer-based spaced repetition system and the direct ancestor of the algorithms powering Anki, FSRS, and modern language learning tools including Sakubo.

In-Depth Explanation

Wozniak’s contribution to learning technology is singular: he transformed spaced repetition from a physical system requiring manual scheduling (as in the Leitner System) into an automated, algorithmic process that could run on a personal computer. This made it possible for learners to maintain thousands of items in an SRS simultaneously, with the computer handling all scheduling — something impossible with physical flashcard boxes.

Beginning in 1982, Wozniak became obsessed with optimizing his own learning efficiency. Frustrated with traditional study methods, he began tracking his learning in meticulous detail — initially with hand-written schedules, then with spreadsheets once he gained access to a computer. He was particularly interested in finding the optimal time to review information: not too soon (wasting time on material not yet forgotten) and not too late (after it had already decayed). His research was self-directed and self-funded, driven by personal curiosity rather than academic requirements.

By 1985, Wozniak had formalized his scheduling model into the first version of SuperMemo, running on a rudimentary personal computer. The core insight was to model the forgetting curve and use it to predict when each item would be forgotten, then schedule a review just before that point. Each successful review pushed the next review further into the future; failed reviews reset the interval. This approach — using individual performance data to personalize review schedules — was novel and powerful.

The SM-2 algorithm, developed in 1987, refined this model with an “easiness factor” (EF) per card that adjusted based on performance, and a clean mathematical formula for calculating intervals. SM-2’s openness — Wozniak published the formula and documentation freely in 1990 — allowed other developers to implement it. Damien Elmes adopted SM-2 as the basis for Anki in 2006, bringing Wozniak’s algorithm to millions of users worldwide — most of whom have never heard his name.

Wozniak has also published extensively on the theory of memory and learning, including his “two-component model of long-term memory” (distinguishing memory retrievability from stability), research on sleep and memory consolidation, and his concept of “incremental reading.” His website, supermemo.com, is one of the most extensive public resources on memory science and SRS algorithm design.

History

• 1982–1985: Wozniak, a student in Poland, begins recording his own learning in meticulous hand-written schedules. Lacking access to scientific journals, he derives his scheduling rules empirically from his own data. By 1985, he has built a computer program — the first version of what would become SuperMemo — running on an Amstrad CPC. [Wozniak, 1990]

• 1987: Develops the SM-2 algorithm, a refinement of his earlier models, using an “easiness factor” that adapts intervals per card based on recall performance. SM-2 is notable for its simplicity and computational efficiency. This version becomes the most widely adopted iteration of Wozniak’s work and forms the basis of most modern SRS tools. [Wozniak, 1990]

• 1991: Founds SuperMemo World Sp. z o.o. in Poland with Krzysztof Biedalak to develop and distribute SuperMemo commercially. The company continues to operate to the present day, maintaining the desktop application and publishing Wozniak’s ongoing research.

• 1994: Co-publishes with Gorzelanczyk the first peer-reviewed academic paper validating the SRS approach, establishing the scientific credibility of algorithmic spaced repetition beyond self-experimentation. [Wozniak & Gorzelanczyk, 1994]

• 1990s–2000s: Wozniak continues developing SuperMemo through versions SM-5, SM-8, SM-11, SM-15, and beyond — each incorporating more sophisticated memory models, including neural network-based scheduling. These later algorithms are used only in SuperMemo; the broader community continues with SM-2 due to its open documentation and simplicity.

• 2006: Damien Elmes releases Anki using SM-2, bringing Wozniak’s algorithm to millions. Anki’s open-source nature and cross-platform availability make it the world’s most widely used SRS — far exceeding SuperMemo’s user base — though both trace their roots to Wozniak’s foundational research.

• 2022: The FSRS algorithm, developed by Jarrett Ye, explicitly extends Wozniak’s two-component memory model using modern machine learning. FSRS’s adoption in Anki represents the next chapter in the research lineage Wozniak began.

• Present: Wozniak continues to maintain supermemo.com, publish research on memory and sleep, and develop SuperMemo. He is considered one of the most important figures in educational technology history, despite remaining relatively unknown outside SRS and language learning communities.

Criticisms

Wozniak’s work on SuperMemo and the SM-2 algorithm has been criticized on several fronts. The SM-2 algorithm uses a single “ease factor” per item, which researchers have argued is too simplistic to model the complexity of human memory — it does not account for item similarity, interference effects, or the context-dependency of retrieval. The newer FSRS algorithm, developed by Jarrett Ye and now integrated into Anki, has demonstrated measurably superior retention prediction in large-scale studies, suggesting SM-2’s mathematical model was a useful approximation but not an optimal one.

Wozniak’s approach to learning optimization has also been criticized for its reductionism: treating knowledge as discrete, schedulable items may undervalue the role of contextual, associative, and narrative encoding that occurs during extensive reading and naturalistic immersion. His advocacy for “incremental reading” within SuperMemo has not produced peer-reviewed evidence of superiority over conventional reading strategies. Finally, SuperMemo’s proprietary, Windows-only software model — in contrast to Anki‘s open-source approach — has been criticized for limiting the broader SRS research community’s ability to replicate and extend his findings.

Related Terms

• SuperMemo
• SM-2 (SuperMemo 2)
• SRS (Spaced Repetition System)
• FSRS (Free Spaced Repetition Scheduler)
• Anki

See Also

• Sebastian Leitner
• Hermann Ebbinghaus
• Forgetting curve
• Spacing effect

Research

• Wozniak, P.A. (1990). Optimization of learning [Master’s thesis]. University of Technology, Poznan. https://www.supermemo.com/en/archives1990-2015/english/ol
  Summary: Wozniak’s original presentation of SuperMemo and the derivation of SM-2. The primary technical reference for the SM-2 algorithm, documenting his empirical approach to deriving scheduling rules from personal learning data.

• Wozniak, P.A., & Gorzelanczyk, E.J. (1994). Optimization of repetition spacing in the practice of learning. Acta Neurobiologiae Experimentalis, 54(1), 59–62.
  Summary: The first peer-reviewed academic paper validating Wozniak’s SRS approach. Demonstrates statistically significant retention advantages for algorithmically scheduled repetition — moving SRS from self-experimentation to scientific validation.

• Wozniak, P.A. (1995). SuperMemo 2: Algorithm. supermemo.com. https://www.supermemo.com/en/blog/application-of-a-computer-to-improve-the-results-obtained-in-working-with-the-leitner-system-of-flashcards
  Summary: Wozniak’s open documentation of the SM-2 algorithm. This is the source that enabled Anki and many other tools to implement SM-2, making it the most widely deployed SRS algorithm in history.

• Ye, J. et al. (2022). FSRS: A stochastic shortest path algorithm for optimizing spaced repetition scheduling. [GitHub: open-spaced-repetition/fsrs4anki]. https://github.com/open-spaced-repetition/fsrs4anki
  Summary: The foundational paper for FSRS, which explicitly builds on and improves Wozniak’s two-component memory model. Provides context for how far SRS algorithm research has advanced since SM-2 and situates Wozniak’s contribution within the broader history of the field.

Note:

• Piotr Wozniak should not be confused with Steve Wozniak, co-founder of Apple. The two are unrelated.
• Wozniak’s later SuperMemo algorithms (SM-5 and above) are proprietary and not publicly documented in the same detail as SM-2. The open publication of SM-2 is one of the key reasons it became so widely influential.