A Thousand Brains presents Jeff Hawkins’s theory that intelligence arises from the neocortex building many maplike models of the world through cortical columns, reference frames, prediction, and sensorimotor learning.
In A Thousand Brains: A New Theory of Intelligence, Jeff Hawkins argues that the key to understanding intelligence lies in the neocortex, the brain structure responsible for perception, language, reasoning, and high-level thought. Building on Vernon Mountcastle’s idea that the neocortex uses a common cortical algorithm, Hawkins proposes that the brain is made of many cortical columns, each capable of learning models of objects and concepts through movement, prediction, and reference frames. The book’s central idea, the Thousand Brains Theory of Intelligence, claims that intelligence does not come from one centralized model of the world. Instead, thousands of cortical columns each build their own partial models, then coordinate through a kind of voting process to produce coherent perception and thought. Hawkins extends this theory from physical perception to language, abstraction, consciousness, artificial intelligence, and humanity’s long-term future. The second half of the book explores what this theory means for AI. Hawkins argues that current AI systems are powerful but not truly intelligent in the biological sense because they lack grounded world models, embodiment, continual learning, and sensorimotor understanding. He also challenges common fears about machine intelligence, arguing that intelligence itself does not automatically create dangerous goals; those drives come from older brain systems, not from the neocortex. The final section turns to human intelligence, false beliefs, existential risk, and knowledge preservation. Hawkins suggests that humanity’s greatest dangers come from old-brain motivations combined with powerful technology and mistaken models of reality. He closes by arguing that preserving knowledge may be humanity’s most important long-term legacy.