Learning Curve The Case for Memorization in the AI Era
13 snips
Mar 3, 2026 
Barbara Oakley, learning-science professor urging a return to internalized facts for real expertise. Nelson Dellis, memory champion and coach who trains vivid mnemonic systems. They debate why rote facts matter in an AI world. They explore timelines, dates, language memorization, mnemonic encoding of numbers, and how practice turns tricks into fluency.
AI Snips
Chapters
Books
Transcript
Episode notes
Inside A Memory Championship
- Nelson Dellis describes competing at the USA Memory Championships where he memorized 104 cards and read them back on stage.
- He trains with vivid mnemonic images and practices until the images become an automatic “movie” he can replay to decode card order.
Memory Paradox And The Reverse Flynn Effect
- Barbara Oakley connects declines in some IQ measures to reduced emphasis on memorization since the 1970s and the rise of calculators.
- She argues internalized knowledge builds schemas that enable transfer and deeper understanding, which discovery-only teaching neglects.
Require Core Facts Before Outsourcing To AI
- Oakley recommends reintroducing explicit memorization for core facts in subjects like math, physics, and history and testing away from AI.
- She suggests mixing memorization with learning to use AI, insisting some things must be internalized first.
