## Overview Learning has convex returns from repetition: reading the same text twice produces more learning than reading two different texts once. This applies Nassim Nicholas Taleb's concept of convexity to knowledge acquisition. The second pass through familiar material yields disproportionately more understanding because the foundational scaffolding is already in place, allowing deeper processing, pattern recognition, and integration with existing knowledge. This principle challenges the common bias toward breadth — the impulse to consume more sources rather than deeply engage with fewer. Volume-based reading creates shallow familiarity across many texts, while depth-based re-reading creates durable understanding and genuine insight from fewer sources. ## Cross-Domain Applications - **Skill Development**: Practicing the same kata, drill, or exercise repeatedly builds deeper automaticity than rotating through many different exercises — the same convexity principle applies to motor learning - **Software Engineering**: Re-reading a codebase or re-implementing a familiar pattern yields deeper architectural understanding than surveying many codebases superficially - **Investment & Decision-Making**: Studying the same company or market deeply (Buffett-style concentration) produces better judgment than scanning many opportunities shallowly