Claims

A cortical column needs only about 100 neurons because that is enough for input/output neurons, a dozen or so AND-gates, and supporting functions, which implies most of the brain's neurons are not there for brute-force learning but are scaffolding waiting to be given meaning through experience.

The neocortex is composed of millions of nearly uniform repeating structures called cortical columns that all start with the same circuitry, and they specialize (e.g., visual cortex tuning to edges/colors, prefrontal cortex to planning) based on the signals they receive, implying the cortex is built from a single repeated general-purpose module.

The brain works not as a statistical engine but as a biological graph processor where concepts are nodes and relationships are precise structured connections, which is the only plausible way the brain could store knowledge.

Each cortical column functions like a node in a graph, with connections representing relationships (is-a, has-a, part-of) that support inheritance, exceptions, and bidirectional links, giving the brain meaningful symbolic reasoning that today's artificial neural networks cannot do.

Learning a relationship in the brain occurs by selectively strengthening a tiny subset of pre-existing near-zero-weight synapses (e.g., from the 'is-a' column to an AND-gate in the Fido column and from that gate to the 'dog' output neuron), so the physical structure is already in place and learning just activates it.

The brain has three major parts: the brain stem managing vital functions like heartbeat and breathing, the cerebellum handling coordination and motor learning, and the neocortex where conscious thought, planning, and reasoning happen.

The brain runs on just 20 watts yet shows understanding and common sense far beyond today's AI systems.

Subject-representation can be interrupted for the better

Just as neurological processes representing the world or body can break down (blindness, motion blindness, inability to name living things), the brain's representation of a subject internal to the body can be interrupted—and unlike most disruptions this one is for the better, halting the usual cascade of psychological suffering and revealing that thoughts arise on their own with no thinker authoring them.

Brain Has a Status Detection System

Neuroscientists describe a 'status detection system' in the brain that constantly measures cues like eye contact with numerical precision; in one study, people served slightly less orange juice than others became preoccupied and upset—not about the juice but because the system read it as a status insult.

The brain computes features rather than detecting them

Sensory experiences like the color red are features the brain computes, not properties it passively detects; seeing red requires combining information from three types of cones registering different wavelength ranges, and the brain similarly computes features like faces, threat, and novelty in the service of regulating the body and action.

Birds lack cortex but share its neurons

Birds do not have a cerebral cortex yet possess neurons that are the same as those making up the mammalian cerebral cortex and that perform similar functions, further undermining the idea that the cortex is a uniquely new 'higher' structure.

Triune brain model contradicts the evidence

The triune brain model—lizard brain for instincts, limbic system for emotion, neocortex for rationality, with reason battling an inner beast—does not match the best scientific evidence; it reflects cultural beliefs about morality and the self rather than actual brain organization.

Decision-making is the preparation for action itself

The intuitive sequence of sense → evaluate → decide → act is wrong; the brain instead projects itself into the future to prepare this action versus that action, and that preparation is the decision, after which it predicts the sensory consequences of those movements—so there is no separate deciding step that precedes acting.

Cortisol signals metabolic outlay, not stress per se

Cortisol is not a stress hormone but a hormone released when the brain predicts a large metabolic outlay; stress is just a big anticipated metabolic expenditure, which can be 'good' or 'bad' depending on its purpose and whether the spent resources are repaid—so the same arousal can be reframed as determination rather than anxiety.

Brain imaging blobs are curated, not raw

The 'lit-up region' images from fMRI are not raw outputs but are curated by scientists through analytic choices; underpowered designs that expect islands of activity produce islands, whereas a study running 400 trials per subject and modeling responses differently found that about 85% of the brain shows increased activity for a simple decision.

Perception is an inverse problem solved by prediction

The brain only ever receives the outcomes of world events, not the events themselves, making perception an inverse problem; to interpret ambiguous sense data (a loud bang, a tug in the chest) the brain must guess using past experience reinstated in its own wiring, and it does this predictively rather than reactively because that is metabolically cheaper.

Lesion deficits are complicated by fibers and suppression

The classic inference that a focal lesion proves a region 'does' a function is complicated because damage may destroy fibers of passage rather than the critical neurons, and lost function can sometimes be recovered by a second lesion—e.g., ablating primary visual cortex causes blindness, but a further lesion in the superior colliculus can restore some vision because the colliculus normally suppresses retinal information reaching V1.

Brain's primary job is body regulation

The brain's most important job is not thinking, seeing, or feeling but regulating the internal systems of the body (heart, lungs, immune, endocrine); thinking, deciding, seeing, and feeling are all performed in the service of bodily regulation.

Brain costs 20% of metabolic budget

The brain is an extremely expensive organ, costing roughly 20% of the body's entire metabolic budget, which makes the question of why brains evolved a meaningful evolutionary puzzle.

Mammal brains share one developmental plan

All mammal brains studied follow the same developmental plan with no new neuron types across species; what differs is the duration of each developmental stage, so larger brains have more of certain neuron types and appear reorganized but contain no genuinely new parts.

Brain networks overlap at rich club hubs

When networks are parsed by treating whatever replicates across subjects as signal, the well-known brain networks are not disconnected Lego-like blocks but overlap, and they overlap in densely connected 'rich club hubs' that coordinate activity across the whole brain and serve as the backbone of neural communication.

Larger brain regions gain abstraction via fewer neurons

In large-brained animals certain cortical regions evolve and develop not by adding neurons but by developing fewer, larger neurons with much more connectivity, which lets the brain summarize information more efficiently and perform abstraction—finding functional similarities among things that look, feel, smell, and taste different.

Brain has no modular mental organs

The brain is not organized into discrete object-like parts (an amygdala that 'does' emotion, basal ganglia that 'do' movement, a prefrontal cortex that 'does' rationality); it is better understood as a single dynamically fluctuating network of neurons defined by their relationships and the features they compute, more like shifting weather patterns than batons passed between regions.

Memory is reassembling the past, not retrieving files

Brains do not store memories and call them up like files in a drawer; remembering is reassembling the past in the present by reinstating prior neural patterns for the purpose of making sense of incoming sense data.

Blinded V1 neurons fire from motor predictions

When rats' retinas are destroyed, primary visual cortex neurons quiet down but ramp back to near-normal firing within 24 hours, with patterns resembling sighted activity; this is likely driven by the anterior cingulate cortex—a primary regulator of the body's internal systems—reinstating visual predictions from past experience, illustrating that perception is generated by motor/regulatory predictions rather than incoming light alone.

Brain distinguishes self-caused from externally-caused sensation

To move and survive, brains must distinguish sense data arising from their own movements (largely predicted and cancelled out) from sense data arising from other sources (the genuinely surprising stuff); if you experience surprise from your own movements, something is seriously wrong, as Barrett discovered when an inner-ear infection destroyed her balance and made ordinary head movements produce vivid, disorienting prediction error.

Motor action is primary; sensing serves movement

Motor action is primary, with all sensory processing in its service, and there is no external movement of bodies without internal systems to support it; in vertebrates the neurons controlling internal viscera (heartbeat, lungs) and those controlling voluntary skeletal muscles are intertwined in the anatomy, even though motor neuroscience tends to ignore the internal systems.

Von Economo neurons may not be a special class

Von Economo neurons are contentious; some anatomists argue they are not a special class but simply very large pyramidal cells that appear in large-brained animals because neurons can scale up with brain size, so their existence does not establish a unique higher-order neuron type.

Brain may be irreducibly complex with no simplification

Understanding means having a compressed shorthand version of complexity such that you no longer need the details; but the brain may be a case where there is no simplification—it is its own most concise description—because if a simpler version were possible, the brain would have evolved to be simpler.

Convolutional neural nets derive from Hubel and Wiesel

Modern convolutional neural networks are based on the original studies of Nobel laureates Hubel and Wiesel on how visual information of the world is broken down into very small features in the visual cortex and then reassembled to produce perception.

Neuron as a selfish single-celled organism wanting to survive

A neuron can be understood as a single-celled organism with its own will to survive, living in the 'weird pond' of your head, doing everything for its own survival; learned wiring diagrams emerge from neurons doing what keeps them alive, and humans are simply the embodiment of all their neurons—we eat chocolate chip cookies because our neurons want glucose.

Brain neurons comparable in number to galaxy stars

There are about 85-86 billion neurons in a normal human brain, and each neuron is itself quite complicated, making the brain the most complex system we know of in the universe.

Real action is in connections, not individual neurons

The crucial part of understanding the brain is the wiring diagram—how neurons are connected to each other, where each gets input from and sends output to—rather than just the individual neurons, because this connectivity (the connectome) is crucially important for learning and memory.

Each neuron connects to ~10,000 others both ways

Each of the 86 billion neurons receives input from on the order of 10,000 different neurons and sends output to 10,000 yet different neurons, which is why describing the brain may be possible but understanding it is another matter.

Brain is sensory-in, motor-out processing

The fundamental purpose of the nervous system is sensory-in, motor-out: it takes information from sense organs, processes it through interconnected neurons, and turns it into a motor reaction; humans add the ability to act on stored, learned information through rumination even absent external stimulation.

Neurons fire based on threshold relative to noise

A neuron has a threshold and fires only when input exceeds the background noise at a given moment—analogous to how a small noise gets attention at night but not during a loud day—and salient signals can be amplified and propagated to all of the neuron's target cells.

Brain size relative to body, not absolute, matters

Humans are smart not because we have the largest brains—elephants and whales have much bigger brains—but because we are highly encephalized (large brain relative to body size) and have a large amount of association cortex, more than any other animal.

Corvids have more neurons than humans

Crows and other corvids actually have more neurons than humans do, but packed into a very small space, demonstrating that absolute neuron count is necessary but not sufficient for human-level intelligence.

Neurons encode signal strength via AM/FM modulation

Neurons differ in their firing behavior—some fire a single action potential, others fire bursts whose frequency encodes signal strength—so input amplitude (AM) is converted into output frequency (FM), and these responses are governed by specific membrane ion channels making the system extremely nonlinear and hard to model.

A cubic millimeter of human brain is 1.4 petabytes

Lichtman's lab mapped a cubic millimeter of human brain—about a millionth of a human brain—which produced 1,400 terabytes (1.4 petabytes) of data, with all the segmented wires made available online via Google.

Learning changes both wiring and synaptic strength

Learned behaviors in mammals and humans come from experience rather than genetic programming, and this works through both changes in the wiring diagram (new connections) and changes in the sensitivity/strength of existing synapses, since evolution exploits any mechanism that is useful.

Cajal inferred brain circuits from sparse Golgi stain

Modern neuroscience began with Ramón y Cajal, who used Golgi's stain that randomly labeled only about 1% of nerve cells; the random sparseness allowed him to see the complete connectivity of individual cells and infer that the brain is made of directional circuits with input on dendrites and output via axons.

Real wiring is infinitely more complex than stick figures

The classic Cajalian stick-figure diagrams of neuron connectivity, based on sparse labeling, led people to assume that was the brain; but when you actually image everything, the wiring is infinitely more complicated than the stick figures suggested.

Nervous systems self-organize with built-in chance variability

In both worms and mammals, the nervous system self-organizes to work but leaves a certain amount to chance—mouse muscle innervation shows the same rule-based range of axon sizes across animals, but the specific location each axon goes to is variable—because anything that mattered for survival would have been highly constrained, so variability marks what doesn't matter.

Brainbow labels every cell a different color

The brainbow technique solves the limitation of the sparse Golgi stain by labeling every neuron a different color, so densely-packed cells can be individually traced; however, in the brain itself the wires are so densely packed that diffraction-limited fluorescence microscopy lacks the resolution to trace them, useful mainly in the peripheral nervous system.

Whole mouse brain would be an exabyte of data

A consortium of labs, with Google's help, is building tools for a proof-of-principle whole-mouse-brain connectome of about 1000 petabytes (a million terabytes, an exabyte); they could start within about five years and finish imaging within about another five years.

Memories may be incompressible like genes

There are probably some rules of connectivity (motifs) in the brain, but memories—being unique to each person's experiences—may be incompressible, much like genes which are each unique because they evolved for a particular purpose; turning an experience into a stable, learned, unique wiring diagram is a genuinely hard problem where we have not reached first base despite long effort.

Understanding the brain is like understanding New York City

No human being could hold the amount of information in a brain; the best we may do is describe what is there in complete detail, which is very different from being able to predict behavior—analogous to asking whether you 'understand' New York City, where so much happens simultaneously that the question barely makes sense, and the brain is far more complex than New York City.

Jennifer Aniston neuron is part of a distributed circuit

In awake epilepsy-surgery patients, a recorded cell in the face-processing region fired specifically for Jennifer Aniston's picture (not for Halle Berry, Brad Pitt, or Aniston-with-Pitt), and a separate cell responded to both Bill Clinton's picture and his signature—but no single cell is in charge; concept recognition is encoded by a distributed circuit of cells, and each cell, receiving 10,000 inputs, participates in millions of circuits.

Cells participate in millions of circuits, frustrating mapping

Because each cell receives 10,000 inputs and the brain must store everything you know and every behavior you can do, every cell participates in many different circuits; we lack a full description of any animal's behavioral repertoire, so a wiring diagram is far more complicated than the single phenomenon of interest, making it hard to isolate the circuit for any one thing.