← All topics
🎓

AI Learning Platform

Notes on building AI-powered learning experiences.

30 posts · AI & LLMs

  1. 1

    How Learning Works: The Science of Teaching and AI Tutors

    A plain-English guide to how people actually learn - memory, cognitive load, retrieval practice - and how to build AI tutors that genuinely teach.
  2. 2

    Why AI Tutors Aren't Real Teachers (Yet)

    AI chatbots can explain, quiz, and summarize beautifully, so why aren't they real teachers? Discover the gap between an AI tutor and an answer machine.
  3. 3

    What Instructional Design Actually Means (Plain English)

    Instructional design is the difference between knowing a subject and teaching it. Here is what it means, in plain English, and why content is not a course.
  4. 4

    How Human Memory Works: Why You Forget and How to Fix It

    How human memory works in plain English: the four-item limit, chunking, schemas, and the forgetting curve, plus simple ways to make learning stick.
  5. 5

    Knowing vs Understanding: Why You Forget a Month Later

    Why finishing a course doesn't mean you learned it. Knowing vs understanding vs remembering a month later, and how to actually retain what you study.
  6. 6

    Cognitive Load Theory: Why Too Much at Once Fails

    Cognitive load theory explains why lessons overwhelm you and slip away. Learn the 4-chunk limit and how to design learning that actually sticks.
  7. 7

    Retrieval Practice: Why Testing Beats Re-reading

    Re-reading feels productive but barely sticks. Retrieval practice and the testing effect explain why pulling facts from memory builds lasting learning.
  8. 8

    Spaced Repetition: How to Beat the Forgetting Curve

    Most of what you learn today is gone within a day. Spaced repetition uses the forgetting curve to make memories stick. Here is how it works and how to use it.
  9. 9

    Why the Study Methods That Feel Best Work Worst

    Interleaving, dual coding, and desirable difficulties: why the study methods that feel hard build lasting memory, and how to use them to actually learn.
  10. 10

    Bloom's Taxonomy: The 6 Levels of Real Understanding

    Bloom's Taxonomy maps six levels of thinking, from recall to creation. Learn the ladder, write sharper learning objectives, and stop faking understanding.
  11. 11

    The 2-Sigma Problem: Why Tutored Students Beat 98%

    Benjamin Bloom found that one-to-one tutoring beats the average classroom by 2 sigma. Here is what the 2-sigma problem means and how AI tutors aim to solve it.
  12. 12

    Zone of Proximal Development: Where Learning Happens

    The Zone of Proximal Development is the sweet spot between too easy and too hard. Learn how scaffolding and worked examples keep learners in it.
  13. 13

    Why Learners Quit Great Apps (And How to Keep Them)

    Even a perfect AI tutor fails if the learner feels bored or judged. Learn how feedback, motivation, and metacognition keep people coming back and learning.
  14. 14

    Why One Explanation Fails (and Layered Lessons Win)

    Learn the layered explanation method that makes ideas stick: simple, deeper, analogy, example, real use, and practice. A clearer way to teach and learn.
  15. 15

    Analogies, Diagrams & Animations: How Ideas Actually Stick

    Learn why analogies, diagrams, animations, and simulations make ideas stick - and the working-memory rule that decides when each one helps or just clutters.
  16. 16

    Why Quizzing Yourself Beats Rereading Every Time

    Retrieval practice and adaptive quizzes build memory that lasts. Learn why testing yourself beats rereading, and how to design quizzes that actually teach.
  17. 17

    The Teach-It-Back Method: Learn Anything Faster

    The teach-it-back method reveals what you really understand. Learn the science behind it, the Feynman Technique, and how AI grades your explanations fairly.
  18. 18

    Build a Tutor Chatbot That Guides, Not Gives Answers

    Learn how a lesson-scoped tutor chatbot keeps learners on track, asks before it tells, and grounds every answer in the actual lesson to prevent wrong info.
  19. 19

    Learner Models: How AI Tutors Remember What You Know

    A learner model is the memory that turns a chatbot into a real tutor. See how knowledge tracing tracks what a student knows and powers true personalization.
  20. 20

    How AI Tutors Decide What to Teach You Next

    A knowledge graph maps which concepts must come first so an AI tutor can build a personal learning path. Here is how that works and where humans must check it.
  21. 21

    What Should a Learner See Next? The Art of Sequencing

    How adaptive tutors decide what a learner sees next: new concepts, spaced review, or remediation. A plain guide to sequencing, the ZPD, and when to review.
  22. 22

    Spaced Repetition Algorithms: From Leitner to SM-2 and FSRS

    How spaced repetition algorithms decide when you review a flashcard next. A plain-language tour of Leitner boxes, SM-2, Anki, and FSRS, and which to use.
  23. 23

    Misconception or Just Forgetting? How AI Tutors Fix Weak Spots

    Learn how AI tutors tell a real misconception from simple forgetting, trace mistakes to their root cause, and repair weak areas so they actually stay fixed.
  24. 24

    Where LLMs Fit in a Tutor - and Where They Quietly Fail

    Learn where an LLM is brilliant for tutoring and where it fails. A plain guide to hallucination, sycophancy, and the orchestration layer that fixes them.
  25. 25

    Turn Any PDF Into a Course: How RAG Really Works

    Learn how RAG turns a 300-page PDF into a tutor that teaches your exact material, cites its sources, and stops making confident mistakes. Plain-English guide.
  26. 26

    How to Stop Your AI Tutor From Confidently Lying

    Learn how to keep an AI tutor accurate and pedagogically sound with grounding, verification, Socratic prompts, and LLM-as-judge so it teaches instead of cheats.
  27. 27

    How to Measure Real Learning (Not Vanity Metrics)

    Lessons completed and streaks look great, but do they prove learning? Learn how to measure real learning with transfer, retention, and metrics that matter.
  28. 28

    Why "Teach Everything" Kills Your Learning App

    An AI tutor that teaches everything competes with free chatbots and loses. Learn why picking a niche makes your learning app smarter and harder to copy.
  29. 29

    AI Tutor Business Model: How to Build a Real Moat

    Most AI tutors are thin wrappers that get copied overnight. Learn the business model and the moat that make an AI learning product genuinely defensible.
  30. 30

    Why Your AI Tutor Must Promise Memory, Not Just Chat

    An AI tutor that only talks has promised nothing. Learn how to build trust, durable memory, and a learning product people actually keep paying for.