%> A Thousand Brains — Jeff Hawkins | HourLife

Jeff Hawkins · 2021 · Neuroscience + AI

A Thousand Brains

Intelligence is not a single voice in your head. It is a parliament of cortical columns, each building its own model and voting on what is real.

Core move: improve prediction quality by creating stronger consensus across many models, not by trusting one fast guess.

150K+
cortical columns
Reference
frames everywhere
Voting
drives decisions

Core Idea

Your Brain Is a Society of Models

Hawkins argues that every cortical column learns complete object models in its own reference frame. Instead of a single top-down controller, the brain is massively parallel: many models predict what comes next, and behavior emerges from consensus.

Practical implication: when you feel uncertain, your goal is not to force certainty. Your goal is to gather better evidence and let more internal models align.

1

Reference Frames

Knowledge is grounded in location and movement. The brain learns not just what something is, but where it is and how it changes with action.

2

Many Parallel Models

Thousands of columns model the world independently. Robust intelligence comes from diversity and redundancy, not one fragile representation.

3

Voting and Consensus

Columns vote on interpretations. Confidence grows when agreement rises and sensory conflict falls. This is the engine of perception and thought.

Interactive Lab

Cortical Consensus Simulator

Tune consensus, sensory noise, and active exploration to see how prediction confidence changes inside a thousand-column model.

Scenario

Kitchen Mug - Familiar object, low ambiguity, stable shape.

72%
24%
38%

Agreeing Columns (of 1,000)

612

Prediction Confidence

76%

Probable match

A working prediction emerges, but dissenting columns still exist.

Next step: add one more sensory check (touch, motion, angle).

Concept Anatomy

How a Single Perception Is Built

Step 1

Sense

Columns receive partial sensory input from different positions and modalities.

Step 2

Predict

Each column compares the input to its own reference-frame model and predicts next states.

Step 3

Vote

Models compete and cooperate. Agreement strengthens one interpretation over alternatives.

Step 4

Act + Update

Movement gathers missing evidence, lowering uncertainty and updating future predictions.

Community Insights

What Readers Keep Highlighting

"The neocortex is composed of many repeated units, each capable of learning complete models of the world."

resonated with this

"Intelligence emerges when many models vote and settle on the most coherent interpretation."

resonated with this

"Reference frames are the foundation of knowledge: knowing what something is depends on knowing where it is."

resonated with this

"Prediction is the cortex's core operation; sensation is interpreted through expected next states."

resonated with this

"Our brains create models of the world, not just reactions to stimuli."

resonated with this

"False beliefs persist when model voting is isolated from contradictory evidence."

resonated with this

Action Steps

Train Better Models This Week

01

Run a daily model check

When confident about a claim, ask: What evidence would make this model fail? Write one disconfirming signal before you decide.

do this
02

Use three-angle learning

For any concept you are studying, view it from at least three reference frames: visual diagram, plain-language explanation, and practical use case.

do this
03

Add movement before decisions

When stuck, change physical context and gather one new data point. Hawkins-style intelligence improves when action updates prediction.

do this
04

Create a consensus note

For hard choices, write your top 3 internal models (optimistic, cautious, skeptical) and let them vote on the next step.

do this
05

Train uncertainty tolerance

Keep a short log of predictions you were wrong about. Focus on refining models instead of protecting certainty.

do this

"Intelligence is not where one model wins. It is where many models learn to agree."

Inspired by Jeff Hawkins

Back to Library

Take It With You

Downloads & Shareables

Print it, pin it, post it. Ways to take A Thousand Brains off the screen and into the world.

Printable · PDF

Action Checklist

Every action from this page as a printable to-do list with a 7-day tracker.

Download PDF →
Social · Image

Book Summary Card

Shareable 1200×630 card with the book and its top-voted insight. Perfect for social.

Preview →
All Sizes · Gallery

Resource library

Preview and download the summary card plus every quote card in 6 sizes — Instagram feed, Story, Pinterest, YouTube thumbnail, phone wallpaper, and OG share.

Quote cards — one per insight
Click to download PNG · hold ⌥ to preview