No. 48 of 500 No. 9 in Mind, Brain & Philosophy

A Thousand Brains

A New Theory of Intelligence

By Jeff Hawkins

Published
March 2, 2021
Print length
288 pages
Publisher
Basic Books
Reading level
Intermediate
ISBN-13
978-1541675810
Written for
Readers curious about how brains create intelligence and how that might inform AI design.

Synopsis

Jeff Hawkins, who helped create handheld computing at Palm before founding a neuroscience institute and the research company Numenta, presents his team's account of how the brain produces intelligence. The first part lays out the Thousand Brains Theory: the neocortex is built from many thousands of similar cortical columns, each of which learns complete models of objects and concepts using map-like reference frames related to the grid cells that help animals navigate. Perception emerges as these many models vote to reach a consensus, which Hawkins argues explains why the brain is so flexible and why our experience of the world feels unified.

The second part applies the theory to machines. Hawkins argues that current AI lacks the movement-based, reference-frame learning that brains use, and that genuinely intelligent machines will need it. He separates the neocortex, which models the world, from older brain regions that supply drives and emotions, and concludes that intelligent machines would not develop dangerous goals unless people built such goals into them. The final part turns to humanity's prospects, including false beliefs, existential risks and ways to preserve human knowledge. Richard Dawkins contributes a foreword. Readers gain a clear introduction to a much-debated theory of the cortex and a contrarian view of AI safety.

Scorecard

Editorial score

How it scored

Editorial score: 68.6 / 1005 weighted criteria

  • Influence30%

    21.0/30
  • Authority25%

    18.8/25
  • Practical value20%

    11.0/20
  • Currency15%

    9.8/15
  • Readability10%

    8.0/10
  • Editorial score

    68.6/100

How the rank was computed

Rank No. 48 of 500 comes from a composite of 69.6: observable signals 38.7 of 55 points plus editorial review 30.9 of 45 points.

All 11 inputs
InputPoints
Observable signals38.7/55
Reach (Amazon ratings volume)12.2/16.5
Consensus across best-of lists11.0/13.8
Awards and bestseller evidence8.3/8.3
Reader rating, sample-adjusted6.2/8.3
Amazon sales rank1.0/5.5
Recency0.0/2.8
Editorial review30.9/45
Author authority8.4/11.3
Influence9.5/13.5
Practical value5.0/9.0
Readability3.6/4.5
Currency4.4/6.8

Each criterion is scored from 1 to 10 and weighted, and the points add up to the editorial score. The rank combines observable signals (55 percent) with this score (45 percent). Read the methodology.

Verdict

An accessible, provocative theory of how the neocortex builds models of the world, more persuasive on neuroscience than on AI risk.

Strengths

  • Clear explanation of cortical columns and maps
  • Engaging account of a scientist's research path
  • Bridges neuroscience and AI design questions

Keep in mind

The theory remains contested among neuroscientists, and critics argue its reassuring treatment of AI risk underestimates how harmless-looking goals can produce dangerous behavior.

Key takeaways

  1. The neocortex is made of many similar cortical columns, each learning complete models of the world rather than processing isolated fragments of sensory input.
  2. Brains organize knowledge with reference frames, map-like structures that let us represent objects, places and even abstract ideas through location and movement.
  3. Hawkins argues that today's AI systems lack this structured world modeling and that truly intelligent machines will require brain-like principles.
  4. Intelligence and motivation are separable; Hawkins contends machines become dangerous mainly if humans give them harmful drives, a view many alignment researchers dispute.

Notable facts

About the author

Jeff Hawkins

Jeff Hawkins earned an electrical engineering degree from Cornell University, founded Palm in 1992 and later co-founded Handspring. In 2002 he founded the Redwood Neuroscience Institute, and in 2005 he co-founded Numenta to reverse-engineer the neocortex. He co-authored On Intelligence with Sandra Blakeslee and was elected to the National Academy of Engineering in 2003 for creating the handheld computing paradigm.

Questions

What does the Thousand Brains Theory claim?
Jeff Hawkins proposes that the neocortex contains many thousands of cortical columns, each learning full models of objects and ideas using map-like reference frames. Rather than relying on one central model, the brain runs many in parallel and reaches a perception through a kind of voting. Hawkins and his Numenta colleagues developed the theory over years of research.
What does A Thousand Brains say about AI?
Hawkins argues that today's AI systems, including deep learning, do not learn structured models of the world the way brains do, and that truly intelligent machines will need brain-like principles such as reference frames. He also argues that intelligence alone does not create dangerous motives, a reassuring position that many AI safety researchers consider too optimistic.
Is A Thousand Brains worth reading in 2026?
It remains a clear, enjoyable introduction to one ambitious theory of the cortex and a useful counterpoint to scaling-focused views of AI. Readers should know that it predates the surge of large language models after 2022, so its comments on machine intelligence feel dated, and other researchers still contest its neuroscience claims.

Sources

As an Amazon Associate, Best AI Books earns from qualifying purchases. Best AI Books is sponsored by Iternal.ai, whose founder and CEO, John Byron Hanby IV, wrote The AI Strategy Blueprint, our #1 ranked book. How we rank books.