Cortical columns
This page describes the organizational forces that limit change.  It explains how to overcome them when necessary. 

Power& tradition holding back progress
This page uses an example to illustrate how:
  • A business can gain focus from targeting key customers,
  • Business planning activities performed by the whole organization can build awareness, empowerment and coherence. 
  • A program approach can ensure strategic alignment. 
Be responsive to market dynamics
This page uses the example of HP's printer organization freeing itself from its organizational constraints to sell a printer targeted at the IBM pc user. 
The constraints are described. 
The techniques to overcome them are implied. 
Overcome reactionaries
Primary Navigation

Cortical columns



Summary
Jeff Hawkins develops a theory to explain the systemic logic of the repeating structure of the cortical columns of the neocortex is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 
.  He shows each column supports modeling of objects mapped via internal reference frames is a coordinate system (set of axis) centered on a particular aspect of the situation that describes the location of an object.  It tells you where things are located relative to each other.  And they can detail how to achieve goals such as movements to get from one location to another.  The brain supports many frames of reference, with which, Jeff Hawkins realized, it can model its proximate environment, including for vision: Geons, LIP, superior colliculi; (2009), hearing: gain modulation; & movement planning: PRR; (Jul 2002).  Auditory stimuli are initially coded in a head-centered reference frame.  The motor system codes actions in reference frames that depend on motor effectors.  Eye movements are codes in a reference frame that depends on the difference between current and desired arm position.  It is often necessary to transform the location representation of the sensory stimulus into a representation appropriate for the motor act.  An eye-centered reference frame depends on the location of the eye in the head.  A retinotopic reference frame depends on the retinal location that is activated by a visual stimulus.  Double-saccade tasks show how the location of the second visual target is coded relative to current and desired eye position (eye-centered).  


Hawkins divides the brain logically into the old brain and the neocortex, highlighting the latter's contribution to human intelligence: vision, language, music, math, science, engineering, high-level thinking; the neocortex is doing most of the cognitive is the ability to orchestrate thought and action in accordance with internal goals according to Princeton's Jonathan Cohen. 
interpretation

Hawkins uses his theory to develop minimal criteria for intelligence and to critique current developments in Machine Intelligence.  He sees the need for a general purpose AI solution, inherently conscious being based on the brain's architecture, rather than the current highly specific technologies that demonstrate no intelligence.  He outlines the broad way forward, and concludes machine intelligence will not become an existential risk for humanity. 

But Hawkins is less sanguine about human intelligence, noting the shortsightedness of emotions are low level fast unconscious agents distributed across the brain and body which associate, via the amygdala and rich club hubs, important environmental signals with encoded high speed sensors, and distributed programs of action to model: predict, prioritize guidance signals, select and respond effectively, coherently and rapidly to the initial signal.  The majority of emotion centered brain regions interface to the midbrain through the hypothalamus.  The cerebellum and basal ganglia support the integration of emotion and motor functions, rewarding rhythmic movement.  The most accessible signs of emotions are the hard to control and universal facial expressions.  Emotions provide prioritization for conscious access given that an animal has only one body, but possibly many cells, with which to achieve its highest level goals.  Because of this, base emotions clash with group goals and are disparaged by the powerful.  Pinker notes a set of group selected emotions which he classes as: other-condemning, other-praising, other-suffering and self-conscious emotions.  Evolutionary psychology argues evolution shaped human emotions during the long period of hunter-gatherer existence in the African savanna.  Human emotions are universal and include: Anger, Appreciation of natural beauty, Contempt, Disgust, Embarrassment, Fear, Gratitude, Grief, Guilt, Happiness, Honor, Jealousy, Liking, Love, Moral awe, Rage, Romantic love, Lust for revenge, Passion, Sadness, Self-control, Shame, Sympathy, Surprise; and the sham emotions and distrust induced by reciprocal altruism. 
, the acceptance and impact of false beliefs and the existential risks enabled by intelligence.  He explores for potential solutions to the dilemma: the potential of integrating brains and machines, preserving what we have discovered, prophylactic actions to limit the damage

Finally Hawkins issues a call to arms

Complex adaptive system (CAS) theory provides an organizing framework within which organisms and knowledge emerge.  CAS theory equates cortical columns to CAS action agents. 

A Thousand Brains
In Jeff Hawkins's book 'A Thousand Brains' he outlines his theory of human intelligence enables the achievement of goals in the face of obstacles.  The goals are sub-goals of genes' survival and reproduction and include:
  • Obtaining and eating food
  • Sex
  • Finding and maintaining shelter
  • Fighting for resources - in the preferred hunter-gatherer environment loss of resources was critical while possession was often transient. 
  • Understanding the proximate environment
  • Securing the cooperation of others 
based on Vernon Mountcastle's model of the Neocortex is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 


1 Part 1: A new understanding of the brain
Many people are striving to understand the brain.  It enables us to
Reading and writing present a conundrum.  The reader's brain contains neural networks tuned to reading.  With imaging a written word can be followed as it progresses from the retina through a functional chain that asks: Are these letters? What do they look like? Are they a word? What does it sound like? How is it pronounced? What does it mean?  Dehaene explains the importance of education in tuning the brain's networks for reading as well as good strategies for teaching reading and countering dyslexia.  But he notes the reading networks developed far too recently to have directly evolved.  And Dehaene asks why humans are unique in developing reading and culture. 

He explains the cultural engineering that shaped writing to human vision and the exaptations and neuronal structures that enable and constrain reading and culture. 

Dehaene's arguments show how cellular, whole animal and cultural complex adaptive system (CAS) are related.  We review his explanations in CAS terms and use his insights to link cultural CAS that emerged based on reading and writing with other levels of CAS from which they emerge. 

read a book and write one
.  But the individual neurons, specialized eukaryotic cells include channels which control flows of sodium and potassium ions across the massively extended cell membrane supporting an electro-chemical wave which is then converted into an outgoing chemical signal transmission from synapses which target nearby neuron or muscle cell receptors.  Neurons are supported by glial cells.  Neurons include a:
  • Receptive element - dendrites
  • Transmitting element - axon and synaptic terminals.  The axon may be myelinated, focusing the signals through synaptic transmission, or unmyelinated - where crosstalk is leveraged. 
  • Highly variable DNA schema using transposons. 
that participate in this process can't do this on their own.  Hawkins asserts that globally, tens of thousands of neuroscientists are trying to understand the mystery of how our brain gives rise to human intelligence enables the achievement of goals in the face of obstacles.  The goals are sub-goals of genes' survival and reproduction and include:
  • Obtaining and eating food
  • Sex
  • Finding and maintaining shelter
  • Fighting for resources - in the preferred hunter-gatherer environment loss of resources was critical while possession was often transient. 
  • Understanding the proximate environment
  • Securing the cooperation of others 
.  But while a lot of details have been discovered there is little understanding of how the whole thing works. 

Francis Crick argued in Thinking About the Brain, we need to understand how to arrange the discovered pieces correctly.  This inspired Hawkins mission - to understand the neocortex is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 
as a system.  He intends to research how the neocortex works in enough detail to explain the biology of the brain and build intelligent machines based on those principles.  He asserts "it is responsible for everything we associate with intelligence, from our senses of: vision, touch, hearing[, smell]; to language in all its forms, to abstract thinking such as mathematics and philosophy." 

In 2016 Hawkins identified a framework that explains the basics of how the neocortex works and suggests a new way to think about intelligence.  It's not yet a complete theory of the brain but, as in the case of
Desmond & Moore paint a picture of Charles Darwin's life, expanded from his own highlights:
  • His naughty childhood, 
  • Wasted schooldays,
  • Apprenticeship with Grant,
  • His extramural activities at Cambridge, walks with Henslow, life with FitzRoy on the Beagle,
  • His growing love for science,
  • London: geology, journal and Lyell. 
  • Moving from Gower Street to Down and writing Origin and other books. 
  • He reviewed his position on religion: the long dispute with Emma, his slow collapse of belief - damnation for unbelievers like his father and brother, inward conviction being evolved and unreliable, regretting he had ignored his father's advice; while describing Emma's side of the argument.  He felt happy with his decision to dedicate his life to science.  He closed by asserting after Self & Cross-fertilization his strength will be exhausted.  
Following our summary of their main points, RSS frames the details from the perspective of complex adaptive system (CAS) theory.  Darwin placed evolution within a CAS framework, and built a network of supporters whose complementary skills helped drive the innovation. 
 
Darwin
's
Richard Dawkin's explores how nature has created implementations of designs, without any need for planning or design, through the accumulation of small advantageous changes. 
theory of evolution
, he expects the details will
This page discusses the mechanisms and effects of emergence underpinning any complex adaptive system (CAS).  Physical forces and constraints follow the rules of complexity.  They generate phenomena and support the indirect emergence of epiphenomena.  Flows of epiphenomena interact in events which support the emergence of equilibrium and autonomous entities.  Autonomous entities enable evolution to operate broadening the adjacent possible.  Key research is reviewed. 
emerge
and aggregate is when a number of actions become coordinated and operate together.  In the adaptive web framework's Smiley, codelets become coordinated by their relative position in the deployment cascade.  The cascade's dynamics are dependent on the situation, the operating codelets responses to that situation and the grouping of schematic strings they are associated with.  The aggregate effect is a phenotype the adaptive agent. 
over
Carlo Rovelli resolves the paradox of time. 
Rovelli initially explains that low level physics does not include time:
  • A present that is common throughout the universe does not exist
  • Events are only partially ordered.  The present is localized
  • The difference between past and future is not foundational.  It occurs because of state that through our blurring appears particular to us
  • Time passes at different speeds dependent on where we are and how fast we travel
  • Time's rhythms are due to the gravitational field
  • Our quantized physics shows neither space nor time, just processes transforming physical variables. 
  • Fundamentally there is no time.  The basic equations evolve together with events, not things 
Then he explains how in a physical world without time its perception can emerge:
  • Our familiar time emerges
    • Our interaction with the world is partial, blurred, quantum indeterminate
    • The ignorance determines the existence of thermal time and entropy that quantifies our uncertainty
    • Directionality of time is real but perspectival.  The entropy of the world in relation to us increases with our thermal time.  The growth of entropy distinguishes past from future: resulting in traces and memories
    • Each human is a unified being because: we reflect the world, we formed an image of a unified entity by interacting with our kind, and because of the perspective of memory
    • The variable time: is one of the variables of the gravitational field.  With our scale we don't register quantum fluctuations, making space-time appear determined.  At our speed we don't perceive differences in time of different clocks, so we experience a single time: universal, uniform, ordered; which is helpful to our decisions

time


Hawkins contends that most of what we know about the proximate environment is learned.  To recognize something we feel and look at it.  To achieve goals we first learn how things behave when we interact with them.  Our intelligence is closely linked to the human brain's
The agents in complex adaptive systems (CAS) must model their environment to respond effectively to it.  Evolution's schematic operators and Samuel modeling together support the indirect recording of past successes and their strategic use by the current agent to learn how to succeed in the proximate environment. 
model
of the world.  Hawkins's framework shows how the neocortex uses most of its neurons to create and manipulate hundreds of thousands of reference frames, to structure the model of the world and plan and think.  He calls this the Thousand Brains Theory

Hawkins explains he split the book into three parts:
  1. Describes the steps the Newmenta team took to develop the Thousand Brains theory of reference frames
  2. Explores intelligent machines using the Thousand Brain theory to show current neural networks can't be intelligent and highlights how truely intelligent machines will be used and why they aren't an existential risk in Hawkins's view.  
  3. Human intelligence is critiqued demonstrating that our model of the world can include false beliefs.  For Hawkins this does induce an existential threat to humanity. 
Hawkins started work as an engineer at Intel, but inspired by Thinking About the Brain he explored becoming a neuroscientist.  But it became clear that he would never obtain grants to work on a theory of the neocortex.  It was considered far too uncertain an outcome.  And Hawkins did not want to become an experimentalist.  So he decided to become an entrepreneur, founding: Gridpad, Palm Computing, Treo; to be able to finance his own theoretical neuroscience research program, first at the Redwood Neuroscience Institute and subsequently at Numenta.  The aim was to develop a theory of how the neocortex works and then apply the theory to machine learning and machine intelligence. 

11 Old Brain--New Brain
Hawkins describes details that will help with understanding how the brain creates intelligence:

Three general observations about the neocortex:
  1. Local circuits in the neocortex are complex - every circuit in the neocortex is doing complex processing. 
  2. The Neocortex looks similar everywhere - when comparing across species or looking between regions.  Hawkins admits their are differences but they are relatively small. 
  3. Every part of the neocortex generates movement - Every part of the neocortex that has been investigated includes cells that project to some part of the 'old brain' related to movement.  He concludes the complex circuits are performing sensory-motor tasks. 
Hawkins sees the neocortex as the organ of intelligence enables the achievement of goals in the face of obstacles.  The goals are sub-goals of genes' survival and reproduction and include:
  • Obtaining and eating food
  • Sex
  • Finding and maintaining shelter
  • Fighting for resources - in the preferred hunter-gatherer environment loss of resources was critical while possession was often transient. 
  • Understanding the proximate environment
  • Securing the cooperation of others 


21 Vernon Mountcastle's Big Idea
Johns Hopkins neuroscientist Vernon Mountcastle wrote a contribution for The Mindful Brain that profoundly impacted Hawkins and the Thousand Brains Theory.  Mountcastle asserts that the neocortex is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 
grew to be 70% of the volume of the brain by replicating the same basic circuit many times.  He sees our particularly large neocortex as having more copies than other animals.  Thus every part of the neocortex is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 
applies the same principle and so its contribution to: seeing, touching, language, high-level thought; are the same.  It just depends what the particular cortical column (a group of cells that all respond to the same sensory input area, which is made up of many physically distinct minicolumns) is connected to.  And when talking with Hawkins, he questioned the accuracy of the Hierarchy of Features model of sensory processing

Mountcastle based his view of a universal principle on:

Hawkins sees the proposal as analogous to
Desmond & Moore paint a picture of Charles Darwin's life, expanded from his own highlights:
  • His naughty childhood, 
  • Wasted schooldays,
  • Apprenticeship with Grant,
  • His extramural activities at Cambridge, walks with Henslow, life with FitzRoy on the Beagle,
  • His growing love for science,
  • London: geology, journal and Lyell. 
  • Moving from Gower Street to Down and writing Origin and other books. 
  • He reviewed his position on religion: the long dispute with Emma, his slow collapse of belief - damnation for unbelievers like his father and brother, inward conviction being evolved and unreliable, regretting he had ignored his father's advice; while describing Emma's side of the argument.  He felt happy with his decision to dedicate his life to science.  He closed by asserting after Self & Cross-fertilization his strength will be exhausted.  
Following our summary of their main points, RSS frames the details from the perspective of complex adaptive system (CAS) theory.  Darwin placed evolution within a CAS framework, and built a network of supporters whose complementary skills helped drive the innovation. 
 
Darwin
's algorithm of
This page reviews the implications of selection, variation and heredity in a complex adaptive system (CAS).  The mechanism and its emergence are discussed. 
evolution by variation, heridity and selection
- both identify a common algorithm enabling a huge diversity of results.  Hawkins resolved to explain how cortical columns work and thus validate Mountcastle's idea. 

29 A model of the world in your head
The success of any animal depends on how effectively it responds to its sensory inputs.  Hawkins realized that during his cortical interpretation of the sense data, he particularly notices unexpected variations in what he is sensing.  And he proposed this must be because his neocortex is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 
is already predicting
This page discusses the interdependence of perception and representation in a complex adaptive system (CAS).  Hofstadter and Mitchell's research with Copycat is reviewed.  The bridging of a node from a network of 'well known' percepts to a new representational instance is discussed as it occurs in biochemistry, within the brain, in consciousness and abstractly. 
what it expects to sense
and then notices the discrepancies.  It can predict because past experience has modeled what is normal.  Hawkins stresses the cortical columns are continuously predicting what their inputs will be.  A
Read Montague explores how brains make decisions.  In particular he explains how:
  • Evolution can create indirect abstract models, such as the dopamine system, that allow
  • Life changing real-time decisions to be made, and how
  • Schematic structures provide encodings of computable control structures which operate through and on incomputable, schematically encoded, physically active structures and operationally associated production functions. 
mis-prediction
results in shifting attention is the mutli-faceted capability allowing access to consciousness.  It includes selective attention, vigilance, allocating attention, goal focus, and meta-awareness.  Hawkins notes its operation involves the thalamus. 
to the error and an update to the
The agents in complex adaptive systems (CAS) must model their environment to respond effectively to it.  Evolution's schematic operators and Samuel modeling together support the indirect recording of past successes and their strategic use by the current agent to learn how to succeed in the proximate environment. 
models
Touch, for example, requires the brain to predict where each finger will be, what they will feel: weight, temperature, sounds that are expected; operations which proceed unconsciously while the predictions are matched. 

Hawkins notes that through experience the brain must add all of these models to the framework initially deployed by the
Plans emerge in complex adaptive systems (CAS) to provide the instructions that agents use to perform actions.  The component architecture and structure of the plans is reviewed. 
genetic plan
and then adapted in evolutionary biology is a trait that increased the number of surviving offspring in an organism's ancestral lineage.  Holland argues: complex adaptive systems (CAS) adapt due to the influence of schematic strings on agents.  Evolution indicates fitness when an organism survives and reproduces.  For his genetic algorithm, Holland separated the adaptive process into credit assignment and rule discovery.  He assigned a strength to each of the rules (alternate hypothesis) used by his artificial agents, by credit assignment - each accepted message being paid for by the recipient, increasing the sender agent's rule's strength (implicit modeling) and reducing the recipient's.  When an agent achieved an explicit goal they obtained a final reward.  Rule discovery used the genetic algorithm to select strong rule schemas from a pair of agents to be included in the next generation, with crossing over and mutation applied, and the resulting schematic strategies used to replace weaker schemas.  The crossing over genetic operator is unlikely to break up a short schematic sequence that provides a building block retained because of its 'fitness';  In Deacon's conception of evolution, an adaptation is the realization of a set of constraints on candidate mechanisms, and so long as these constraints are maintained, other features are arbitrary. 
to the proximate environment from conception, during childhood and through adolescence.  The neocortex must
The agents in complex adaptive systems (CAS) must model their environment to respond effectively to it.  Evolution's schematic operators and Samuel modeling together support the indirect recording of past successes and their strategic use by the current agent to learn how to succeed in the proximate environment. 
learn
a huge number of things: all the objects is a collection of: happenings, occurrences and processes; including emergent entities, as required by relativity, explains Rovelli.  But natural selection has improved our fitness by representing this perception, in our minds, as an unchanging thing, as explained by Pinker.  Dehaene explains the object modeling and construction process within the unconscious and conscious brain.  Hawkins elaborates how the cortical columns of the neocortex identify and locate objects in our proximate environment and generate the stable representation.  Mathematicians view anything that can be defined and used in deductive reasoning and mathematical proofs as an object.  These mathematical objects can be values of variables, allowing them to be used in formulas.  
and their attributes and positions, that you are aware of around you.  And all the past experiences and high-level concepts.  All the
Reading and writing present a conundrum.  The reader's brain contains neural networks tuned to reading.  With imaging a written word can be followed as it progresses from the retina through a functional chain that asks: Are these letters? What do they look like? Are they a word? What does it sound like? How is it pronounced? What does it mean?  Dehaene explains the importance of education in tuning the brain's networks for reading as well as good strategies for teaching reading and countering dyslexia.  But he notes the reading networks developed far too recently to have directly evolved.  And Dehaene asks why humans are unique in developing reading and culture. 

He explains the cultural engineering that shaped writing to human vision and the exaptations and neuronal structures that enable and constrain reading and culture. 

Dehaene's arguments show how cellular, whole animal and cultural complex adaptive system (CAS) are related.  We review his explanations in CAS terms and use his insights to link cultural CAS that emerged based on reading and writing with other levels of CAS from which they emerge. 

words we can speak, hear, read and write
.  It is this model that enables our predictions, perceptions are internal appearences of the external world and the body according to Haikonen.  RSS views them as evolved models that are:
  • Associated schematically with the signals generated in response to epi-phenomena detected by sensory receptors and
  • Acted on by emergent agents.  
and
Plans are interpreted and implemented by agents.  This page discusses the properties of agents in a complex adaptive system (CAS). 
It then presents examples of agents in different CAS.  The examples include a computer program where modeling and actions are performed by software agents.  These software agents are aggregates. 
The participation of agents in flows is introduced and some implications of this are outlined. 
actions


The brain learns by moving.  Its inputs are constantly changing as the world changes and we move.  It learns by monitoring its inputs changing over time.  Sounds change over time but to learn about external objects we move them and feel them with our hands and observe them with our eyes.  The sensory changes will depend on how our bodies move and how the object moves.  If the brain's predictions of what changes does not match the sensory experience, the neocortex
The agents in complex adaptive systems (CAS) must model their environment to respond effectively to it.  Evolution's schematic operators and Samuel modeling together support the indirect recording of past successes and their strategic use by the current agent to learn how to succeed in the proximate environment. 
updates that aspect of its model


Hawkins notes that our thoughts, ideas and perceptions are activity of neurons, specialized eukaryotic cells include channels which control flows of sodium and potassium ions across the massively extended cell membrane supporting an electro-chemical wave which is then converted into an outgoing chemical signal transmission from synapses which target nearby neuron or muscle cell receptors.  Neurons are supported by glial cells.  Neurons include a:
  • Receptive element - dendrites
  • Transmitting element - axon and synaptic terminals.  The axon may be myelinated, focusing the signals through synaptic transmission, or unmyelinated - where crosstalk is leveraged. 
  • Highly variable DNA schema using transposons. 
.  And all the stored model details are represented through neuroplasticity is lasting change to the brain that occurs throughout life.  It is also termed neural plasticity.  The changes include:
  • The strength of dendritic input alters due to genetic, neural and hormonal signals
    • Hebb notes that memories require strengthening of preexisting synapses.  Glutamate responsive neurons' post synaptic dendritic spines have two types of receptor: non-NMDA and NMDA.  NMDA channels are responsible for this strengthening mechanism.  LTP then occurs to prolong the increase in excitability of the synapse. 
    • The LTP operation results in calcium diffusion which triggers new spine formation in adjacent parts of the dendrite.  Eventually that can stimulate dentrite growth enabling more neurons to connect. 
    • Short term stress promotes hippocampal LTP.  
    • Sustained stress promotes:
      • Hippocampal & frontal cortex  LTD & suppresses LTP.  Subsequent reductions in NCAM then reduce dendrite and synapse density. 
      • Amygdala LTP and suppresses LTD boosting fear conditioning.  It increases BDNF levels and expands dendrites in the BLA. 
    • Depression and anxiety reduce hippocampal dendrite and spine number by reducing BDNF. 
  • The axon's conditions for
    • Initiating an action potential. 
      • Progesterone boosts GABA-ergic neurons response to GABA decreasing the excitability of other neurons over a period of hours. 
    • Duration of a neuron's refractory period.  Testosterone shortens the refractory period of amygdala and amygdala target neurons over a period of hours. 
  • Synaptic connections being constantly removed and recreated
  • Synapses being created or destroyed.  Stimulation generates additional dendritic spines which become associated with a nearby axon terminal and within weeks a synapse forms.  The synapse then contributes calcium diffusion through LTP triggering more spine formation.  When dendritic spines recede synapses disappear. 
  • Cortical maps change to reflect alterations in the inputs and outputs from the body. 
  • Birth of brain cells in many areas of adult brains: the hippocampus (where 3% are replaced each month) and olfactory bulb and lesser amounts in the cortex. 
  • Restructuring after brain damage including axonal plasticity.  Distant rerouting of axons is observed but no mechanism has been identified yet. 
  • Vision is plastic in predators, where the eyes are moved during final development.  Dehaene argues for neuronal recycling supporting reading.  
rather than just synaptic strengthening, a neuron structure which provides a junction with other neurons.  It generates signal molecules, either excitatory or inhibitory, which are kept in vesicles until the synapse is stimulated when the signal molecules are released across the synaptic cleft from the neuron.  The provisioning of synapses is under genetic control and is part of long term memory formation as identified by Eric Kandel.  Modulation signals (from slow receptors) initiate the synaptic strengthening which occurs in memory. 
.  The connections in the network of neurons represent the details of the world that we have experienced. 

39 The brain reveals its secrets
Hawkins notes the benefit of using the appropriate theoretical framework is the removal of confusion in comprehending complex, M. Mitchell Waldrop describes a vision of complexity via:
  • Rich interactions that allow a system to undergo spontaneous self-organization and, for CAS, evolution
  • Systems that are adaptive
  • More predictability than chaotic systems by bringing order and chaos into
  • Balance at the edge of chaos
situations.  Coherent systemic frameworks have been developed for celestial mechanics,
This page reviews the implications of selection, variation and heredity in a complex adaptive system (CAS).  The mechanism and its emergence are discussed. 
evolution
, genetics, quantum mechanics extends mechanics to atomic and subatomic scales.  Energy, momentum, angular momentum are all restricted to quantized values and all objects have particle and wave properties.  There are various ways to understand quantum mechanics including the cascade of ideas initiated by Hugh Everett. 
and relativity.  Hawkins sees the neocortex is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 
similarly needing a systemic framework that will make experimental observations coherent. 

Hawkins lists the key clarifying leaps in Numenta's research:
  1. The neocortex is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 
    learns a predictive
    The agents in complex adaptive systems (CAS) must model their environment to respond effectively to it.  Evolution's schematic operators and Samuel modeling together support the indirect recording of past successes and their strategic use by the current agent to learn how to succeed in the proximate environment. 
    model
    of the world - but how did it do this?  Hawkins did not see evidence of their being two networks, one making predictions and the other noting what was actually occurring, and it was only when things weren't as expected that he noticed any predictions at all.  Eventually they realized
  2. Predictions occur inside neurons, specialized eukaryotic cells include channels which control flows of sodium and potassium ions across the massively extended cell membrane supporting an electro-chemical wave which is then converted into an outgoing chemical signal transmission from synapses which target nearby neuron or muscle cell receptors.  Neurons are supported by glial cells.  Neurons include a:
    • Receptive element - dendrites
    • Transmitting element - axon and synaptic terminals.  The axon may be myelinated, focusing the signals through synaptic transmission, or unmyelinated - where crosstalk is leveraged. 
    • Highly variable DNA schema using transposons. 
    - Accepting Mountcastle's view of cortical columns and that there are predictions about external objects is a collection of: happenings, occurrences and processes; including emergent entities, as required by relativity, explains Rovelli.  But natural selection has improved our fitness by representing this perception, in our minds, as an unchanging thing, as explained by Pinker.  Dehaene explains the object modeling and construction process within the unconscious and conscious brain.  Hawkins elaborates how the cortical columns of the neocortex identify and locate objects in our proximate environment and generate the stable representation.  Mathematicians view anything that can be defined and used in deductive reasoning and mathematical proofs as an object.  These mathematical objects can be values of variables, allowing them to be used in formulas.  
    and about your movements and how the two interact, suggested to Hawkins that each cortical column must make both types of predictions.  Studying sequence memory provides a model of a sequence of events, such as hearing a sequence of musical notes, which is occurring.  Jeff Hawkins theory of sequence memory asserts its specialized state changes all occur within neurons.  It highlights how the 90% of synapses on dendrites of neurons in a neocortical column are deployed in generating dendritic spikes that raise the potential of the neuron, acting as a prediction of a known sequence that the neuron is participating in modeling.  If new sensory details agree with the prediction the primed neuron will rapidly fire its axon and send inhibitory signals to neurons that are providing alternative sequence models and so are not primed.  If the sensory sequence does not match the sequence memory models then no inhibition from dendritic spikes will occur and lots of neurons will fire indicating a prediction failure. 
    , as it makes predictions about: musical is a complex emergent capability supported by sexual selection and generating pleasure.  It transforms the sensing of epiphenomena: Contour, Rhythm, Tempo, Timbre; to induce salient representations: Harmony, Key, Loudness, Melody, Meter, Pitch, and perceptions: Reverberation - echo; which allow musicians: Elton John, Elvis Presley; to show their fitness: superior coordination, creativity, adolescent leadership, stamina; true for birds and humans.  Levitin showed that listening to music causes a cascade of brain regions to become activated in a particular order: auditory cortex, frontal regions, such as BA44 and BA47, and finally the mesolimbic system, culminating in the nucleus accumbens.  And he found the cerebellum and basal ganglia were active throughout the session.  He argues music mimics some of the features of language and conveys some of the same emotions.  The brain regions pulse with the beat and predict the next one.  As the music is heard it is modeled and generates dopamine rewards for matching each beat and noting creative jokes in the rhythm.  The cerebellum finds pleasure in adjusting itself to stay synchronized. 
    melody is the main, most salient, theme of a piece of music in the mind. 
    , or recognizing a spoken word from a sequence of phonemes are the smallest speech units explicitly representing discrete speech sounds.  The phoneme 't' is a part of Tuna, stop and foot. 
    , or a pattern of often used movements; could highlight the general mechanism.  It took years identifying requirements and
    Terrence Deacon explores how constraints on dynamic flows can induce emergent phenomena which can do real work.  He shows how these phenomena are sustained.  The mechanism enables the development of Darwinian competition. 
    constraints
    , until in 2010 a creative solution became conscious is, argues Stanislas Dehaene, when some attended information eventually enters our awareness and becomes reportable to others.   to Hawkins:
  3. The secret of the cortical column is reference frames is a coordinate system (set of axis) centered on a particular aspect of the situation that describes the location of an object.  It tells you where things are located relative to each other.  And they can detail how to achieve goals such as movements to get from one location to another.  The brain supports many frames of reference, with which, Jeff Hawkins realized, it can model its proximate environment, including for vision: Geons, LIP, superior colliculi; (2009), hearing: gain modulation; & movement planning: PRR; (Jul 2002).  Auditory stimuli are initially coded in a head-centered reference frame.  The motor system codes actions in reference frames that depend on motor effectors.  Eye movements are codes in a reference frame that depends on the difference between current and desired arm position.  It is often necessary to transform the location representation of the sensory stimulus into a representation appropriate for the motor act.  An eye-centered reference frame depends on the location of the eye in the head.  A retinotopic reference frame depends on the retinal location that is activated by a visual stimulus.  Double-saccade tasks show how the location of the second visual target is coded relative to current and desired eye position (eye-centered).  
    - When we move, our cortical columns must model the way we move to be able to make predictions about what will happen next in the proximate context.  Hawkins concluded the sequence-memory provides a model of a sequence of events, such as hearing a sequence of musical notes, which is occurring.  Jeff Hawkins theory of sequence memory asserts its specialized state changes all occur within neurons.  It highlights how the 90% of synapses on dendrites of neurons in a neocortical column are deployed in generating dendritic spikes that raise the potential of the neuron, acting as a prediction of a known sequence that the neuron is participating in modeling.  If new sensory details agree with the prediction the primed neuron will rapidly fire its axon and send inhibitory signals to neurons that are providing alternative sequence models and so are not primed.  If the sensory sequence does not match the sequence memory models then no inhibition from dendritic spikes will occur and lots of neurons will fire indicating a prediction failure. 
    circuit could also make this movement linked prediction if it had an additional input describing how the
    Agents use sensors to detect events in their environment.  This page reviews how these events become signals associated with beneficial responses in a complex adaptive system (CAS).  CAS signals emerge from the Darwinian information model.  Signals can indicate decision summaries and level of uncertainty. 
    sensor
    was moving.  Eventually Hawkins realized the additional input must relate the sensors position relative to the object is a collection of: happenings, occurrences and processes; including emergent entities, as required by relativity, explains Rovelli.  But natural selection has improved our fitness by representing this perception, in our minds, as an unchanging thing, as explained by Pinker.  Dehaene explains the object modeling and construction process within the unconscious and conscious brain.  Hawkins elaborates how the cortical columns of the neocortex identify and locate objects in our proximate environment and generate the stable representation.  Mathematicians view anything that can be defined and used in deductive reasoning and mathematical proofs as an object.  These mathematical objects can be values of variables, allowing them to be used in formulas.  
    , telling him there must be neurons, specialized eukaryotic cells include channels which control flows of sodium and potassium ions across the massively extended cell membrane supporting an electro-chemical wave which is then converted into an outgoing chemical signal transmission from synapses which target nearby neuron or muscle cell receptors.  Neurons are supported by glial cells.  Neurons include a:
    • Receptive element - dendrites
    • Transmitting element - axon and synaptic terminals.  The axon may be myelinated, focusing the signals through synaptic transmission, or unmyelinated - where crosstalk is leveraged. 
    • Highly variable DNA schema using transposons. 
    in the columns representing the location of each sensor in a reference frame 'attached' to the external object.  Each object and each part of the body that is sensing, via touch or vision, must be separately modeled in terms of location and reference frame.  Hawkins realized the neocortex's is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 
    cortical columns were tracking location and modeling reference frames.  And the
    Terrence Deacon explores how constraints on dynamic flows can induce emergent phenomena which can do real work.  He shows how these phenomena are sustained.  The mechanism enables the development of Darwinian competition. 
    constraints
    that he identified as he worked on the theory were all satisfied by his solution to the three leaps. 
Hawkins's
The agents in complex adaptive systems (CAS) must model their environment to respond effectively to it.  Evolution's schematic operators and Samuel modeling together support the indirect recording of past successes and their strategic use by the current agent to learn how to succeed in the proximate environment. 
model
shows how a single cortical column can learn the three dimensional shape of objects by sensing and moving repeatedly.  It leverages a circuit supporting sequence memory, and then adds a layer of neurons identifying location and a second layer representing the object being sensed.  Many thousands of reference frames are active in parallel.  He suggested the reference frames in the cortical columns, are based on an exaptation, initially termed pre-adaptation refers to the coopting of some function for a new use.  It provides an operational phenotype with schematic access to the adjacent possible through the action of regular genetic operators. 
, leveraging the mechanism
This page reviews the implications of selection, variation and heredity in a complex adaptive system (CAS).  The mechanism and its emergence are discussed. 
evolved
within the entorhinal cortex is a main limbic association area between the hippocampus and the neocortex, in the medial temporal lobe.  It is a hub for memory and navigation: location awareness is supported by grid cells.  It is the first brain area impacted in Alzheimer's disease. 


Circuits in the neocortex is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 
learn and predict sequences, and know the location of their input relative to the sensed object by a reference frame fixed to that object, which helps with predicting the next sensory input when we are moving at the same time. 

57 Maps in the brain
Hawkins observes, we perceive are internal appearences of the external world and the body according to Haikonen.  RSS views them as evolved models that are:
  • Associated schematically with the signals generated in response to epi-phenomena detected by sensory receptors and
  • Acted on by emergent agents.  
objects is a collection of: happenings, occurrences and processes; including emergent entities, as required by relativity, explains Rovelli.  But natural selection has improved our fitness by representing this perception, in our minds, as an unchanging thing, as explained by Pinker.  Dehaene explains the object modeling and construction process within the unconscious and conscious brain.  Hawkins elaborates how the cortical columns of the neocortex identify and locate objects in our proximate environment and generate the stable representation.  Mathematicians view anything that can be defined and used in deductive reasoning and mathematical proofs as an object.  These mathematical objects can be values of variables, allowing them to be used in formulas.  
as being somewhere, at a location out in the world, indicating neurons, specialized eukaryotic cells include channels which control flows of sodium and potassium ions across the massively extended cell membrane supporting an electro-chemical wave which is then converted into an outgoing chemical signal transmission from synapses which target nearby neuron or muscle cell receptors.  Neurons are supported by glial cells.  Neurons include a:
  • Receptive element - dendrites
  • Transmitting element - axon and synaptic terminals.  The axon may be myelinated, focusing the signals through synaptic transmission, or unmyelinated - where crosstalk is leveraged. 
  • Highly variable DNA schema using transposons. 
with activity representing the location of all these objects.  This understanding highlights an
This page reviews the implications of selection, variation and heredity in a complex adaptive system (CAS).  The mechanism and its emergence are discussed. 
evolutionary
advantage of knowing where we are in the environment, recollecting: how to get back to the base camp, where fruits and vegetables grow, where a source of water exists.  But to do this requires a reference frame is a coordinate system (set of axis) centered on a particular aspect of the situation that describes the location of an object.  It tells you where things are located relative to each other.  And they can detail how to achieve goals such as movements to get from one location to another.  The brain supports many frames of reference, with which, Jeff Hawkins realized, it can model its proximate environment, including for vision: Geons, LIP, superior colliculi; (2009), hearing: gain modulation; & movement planning: PRR; (Jul 2002).  Auditory stimuli are initially coded in a head-centered reference frame.  The motor system codes actions in reference frames that depend on motor effectors.  Eye movements are codes in a reference frame that depends on the difference between current and desired arm position.  It is often necessary to transform the location representation of the sensory stimulus into a representation appropriate for the motor act.  An eye-centered reference frame depends on the location of the eye in the head.  A retinotopic reference frame depends on the retinal location that is activated by a visual stimulus.  Double-saccade tasks show how the location of the second visual target is coded relative to current and desired eye position (eye-centered).  
, which is provided by "map-creating" neurons in the hippocampus is a part of the medial temporal lobe of the brain involved in the temporary storage or coding of long-term episodic memory.  It includes the dentate gyrus.  Memory formation in the cells of the hippocampus uses the MAP kinase signalling network which is impacted by sleep deprivation.  The hippocampus dependent memory system is directly affected by cholinergic changes throughout the wake-sleep cycle.  Increased acetylcholine during REM sleep promotes information attained during wakefulness to be stored in the hippocampus by suppressing previous excitatory connections while facilitating encoding without interference from previously stored information.  During slow-wave sleep low levels of acetylcholine cause the release of the suppression and allow for spontaneous recovery of hippocampal neurons resulting in memory consolidation.  It was initially associated with memory formation by McGill University's Dr. Brenda Milner, via studies of 'HM' Henry Molaison, whose medial temporal lobes had been surgically destroyed leaving him unable to create new explicit memories.  The size of neurons' dendritic trees expands and contracts over a female rat's ovulatory cycle, with the peak in size and cognitive skills at the estrogen high point.  Adult neurogenesis occurs in the hippocampus (3% of neurons are replaced each month) where the new neurons integrate into preexisting circuits.  It is enhanced by learning, exercise, estrogen, antidepressants, environmental enrichment, and brain injury and inhibited by various stressors explains Sapolsky.  Prolonged stress makes the hippocampus atrophy.  He notes the new neurons are essential for integrating new information into preexisting schemas -- learning that two things you thought were the same are actually different.  Specific cells within the hippocampus and its gateway, the entorhinal cortex, are compromised by Alzheimer's disease.  It directly signals area 25. 
and entorhinal cortex is a main limbic association area between the hippocampus and the neocortex, in the medial temporal lobe.  It is a hub for memory and navigation: location awareness is supported by grid cells.  It is the first brain area impacted in Alzheimer's disease. 
within the limbic system resides mainly in the forebrain, on both sides of the thalamus, it supports emotional circuits: subcortical parts (Amygdala, Septum, Nucleus accumbens), Cortical parts (Hippocampus, Orbitofrontal cortex, Entorhinal cortex, Piriform cortex), Habenula, Diencephalon parts (Mammillary bodies, hypothalamus, anterior nucleus of the thalamus); all of which signal the midbrain through the hypothalamus. The broad interconnections of these regions with a part of the frontal lobe suggested to Walle Nauta that it (ventromedial prefrontal cortex) is a quasi-member of the limbic network.  .  There are place cells are neurons which fire whenever a higher animal occupies a certain location in space.  Place cells are highly invariant over a variety of sensory cues, and they even maintain their space-selective firing as the animal wanders around in full darkness.  They encode where the animal thinks it is.  And the same cells associate a time stamp with the memories.   in the hippocampus that activate when an animal enters a particular location in the proximate environment, and grid cells, are neurons in the entorhinal cortex, support location modeling.  Grid cells activate when a higher animal reaches a particular position in a triangular grid mapping the floor of the proximate space.  Grid cells are lost early in Alzheimer's disease corresponding to a sense of being lost.  Doeller, Barry and Burgess used fMRI with subjects performing a navigation task: moving around a virtual world on a computer screen; to show that grid like cells also operate in the frontal area of the neocortex.  Constantinescu, O'Reilly and Behrens used fMRI to monitor the frontal neocortex of subjects being shown images of birds with different length necks and legs.  The subjects then imagined new birds that combined the neck of one of the birds with the legs of another.  Grid like cells were observed operating, supporting a reference frame map where one dimension represented neck length and another represented leg length.  And the subjects, in thinking about the birds, were mentally moving through the map.  To think about the birds the neocortex created a new map suited to remembering birds with different necks and legs and "movement" was used to recall the details. 
in the entorhinal cortex which actively mark out a grid.  They provide details supporting mapping and recollection of our environment and where our body is in it. 

Hawkins proposed that cortical columns must represent a
The agents in complex adaptive systems (CAS) must model their environment to respond effectively to it.  Evolution's schematic operators and Samuel modeling together support the indirect recording of past successes and their strategic use by the current agent to learn how to succeed in the proximate environment. 
modeled
object is a collection of: happenings, occurrences and processes; including emergent entities, as required by relativity, explains Rovelli.  But natural selection has improved our fitness by representing this perception, in our minds, as an unchanging thing, as explained by Pinker.  Dehaene explains the object modeling and construction process within the unconscious and conscious brain.  Hawkins elaborates how the cortical columns of the neocortex identify and locate objects in our proximate environment and generate the stable representation.  Mathematicians view anything that can be defined and used in deductive reasoning and mathematical proofs as an object.  These mathematical objects can be values of variables, allowing them to be used in formulas.  
's orientation in a similar manner, based on an exaptation, initially termed pre-adaptation refers to the coopting of some function for a new use.  It provides an operational phenotype with schematic access to the adjacent possible through the action of regular genetic operators. 
of the limbic functions.  There are 150,000 copies of this mapping facility, one in each cortical column allowing the neocortex is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 
to track thousands of locations simultaneously including each
Agents use sensors to detect events in their environment.  This page reviews how these events become signals associated with beneficial responses in a complex adaptive system (CAS).  CAS signals emerge from the Darwinian information model.  Signals can indicate decision summaries and level of uncertainty. 
sensory
patch of skin and part of the retina.  And they reference the maplike models in parallel, associatively

65 Maps in a Cortical Column
A layer of neurons (analogous to place cells are neurons which fire whenever a higher animal occupies a certain location in space.  Place cells are highly invariant over a variety of sensory cues, and they even maintain their space-selective firing as the animal wanders around in full darkness.  They encode where the animal thinks it is.  And the same cells associate a time stamp with the memories.  ) respond to sensory inputs when these match their observed features.  A second layer (functioning like grid cells, are neurons in the entorhinal cortex, support location modeling.  Grid cells activate when a higher animal reaches a particular position in a triangular grid mapping the floor of the proximate space.  Grid cells are lost early in Alzheimer's disease corresponding to a sense of being lost.  Doeller, Barry and Burgess used fMRI with subjects performing a navigation task: moving around a virtual world on a computer screen; to show that grid like cells also operate in the frontal area of the neocortex.  Constantinescu, O'Reilly and Behrens used fMRI to monitor the frontal neocortex of subjects being shown images of birds with different length necks and legs.  The subjects then imagined new birds that combined the neck of one of the birds with the legs of another.  Grid like cells were observed operating, supporting a reference frame map where one dimension represented neck length and another represented leg length.  And the subjects, in thinking about the birds, were mentally moving through the map.  To think about the birds the neocortex created a new map suited to remembering birds with different necks and legs and "movement" was used to recall the details. 
) responds to the monitored sensor's movements, identifying a location on a map.  And there are connections associating the two together, so the feature will indicate the location and visa versa.  When movement occurs the location changes causing a prediction of the feature that will be sensed.  When the initial feature is ambiguous all the locations are activated.  Movement will generate multiple predicted sensory details, one or more of which may be matched.  Other locations are eliminated.  Learning a new object is a collection of: happenings, occurrences and processes; including emergent entities, as required by relativity, explains Rovelli.  But natural selection has improved our fitness by representing this perception, in our minds, as an unchanging thing, as explained by Pinker.  Dehaene explains the object modeling and construction process within the unconscious and conscious brain.  Hawkins elaborates how the cortical columns of the neocortex identify and locate objects in our proximate environment and generate the stable representation.  Mathematicians view anything that can be defined and used in deductive reasoning and mathematical proofs as an object.  These mathematical objects can be values of variables, allowing them to be used in formulas.  
includes discovering the associative linkages between the two layers. 

67 Orientation
A column must also represent the sensors orientation.  Head direction cells supports orientation of the head within the brain.  These cells indicate which direction the head is facing with a frame of reference bound to the features of the proximate environment such as the room the animal is in.  And in the neocortex, they provide orientation details relative to landmarks in the proximate external environment.  Head direction cells maintain our sense of direction as our head and body move.  They are present in the entorhinal cortex, retrosplenial cortex, anterior dorsal and lateral dorsal thalamus, lateral mammillary nucleus, dorsal tegmental nucleus and striatum. 
provide this capability for the limbic system resides mainly in the forebrain, on both sides of the thalamus, it supports emotional circuits: subcortical parts (Amygdala, Septum, Nucleus accumbens), Cortical parts (Hippocampus, Orbitofrontal cortex, Entorhinal cortex, Piriform cortex), Habenula, Diencephalon parts (Mammillary bodies, hypothalamus, anterior nucleus of the thalamus); all of which signal the midbrain through the hypothalamus. The broad interconnections of these regions with a part of the frontal lobe suggested to Walle Nauta that it (ventromedial prefrontal cortex) is a quasi-member of the limbic network.   and other brain regions.  Hawkins labels the cortical column equivalents: orientation cells. 

The cortical columns
The agents in complex adaptive systems (CAS) must model their environment to respond effectively to it.  Evolution's schematic operators and Samuel modeling together support the indirect recording of past successes and their strategic use by the current agent to learn how to succeed in the proximate environment. 
learn models
of objects is a collection of: happenings, occurrences and processes; including emergent entities, as required by relativity, explains Rovelli.  But natural selection has improved our fitness by representing this perception, in our minds, as an unchanging thing, as explained by Pinker.  Dehaene explains the object modeling and construction process within the unconscious and conscious brain.  Hawkins elaborates how the cortical columns of the neocortex identify and locate objects in our proximate environment and generate the stable representation.  Mathematicians view anything that can be defined and used in deductive reasoning and mathematical proofs as an object.  These mathematical objects can be values of variables, allowing them to be used in formulas.  
, such as
Tools and the businesses that produce them have evolved dramatically.  W Brian Arthur shows how this occurred.
tools
, via strategies already used by
This page reviews the implications of selection, variation and heredity in a complex adaptive system (CAS).  The mechanism and its emergence are discussed. 
evolutionarily
older limbic brain regions to model the body's proximate environments.  The columns must include exaptation, initially termed pre-adaptation refers to the coopting of some function for a new use.  It provides an operational phenotype with schematic access to the adjacent possible through the action of regular genetic operators. 
and leverage the grid cell, are neurons in the entorhinal cortex, support location modeling.  Grid cells activate when a higher animal reaches a particular position in a triangular grid mapping the floor of the proximate space.  Grid cells are lost early in Alzheimer's disease corresponding to a sense of being lost.  Doeller, Barry and Burgess used fMRI with subjects performing a navigation task: moving around a virtual world on a computer screen; to show that grid like cells also operate in the frontal area of the neocortex.  Constantinescu, O'Reilly and Behrens used fMRI to monitor the frontal neocortex of subjects being shown images of birds with different length necks and legs.  The subjects then imagined new birds that combined the neck of one of the birds with the legs of another.  Grid like cells were observed operating, supporting a reference frame map where one dimension represented neck length and another represented leg length.  And the subjects, in thinking about the birds, were mentally moving through the map.  To think about the birds the neocortex created a new map suited to remembering birds with different necks and legs and "movement" was used to recall the details. 
, place cell are neurons which fire whenever a higher animal occupies a certain location in space.  Place cells are highly invariant over a variety of sensory cues, and they even maintain their space-selective firing as the animal wanders around in full darkness.  They encode where the animal thinks it is.  And the same cells associate a time stamp with the memories.   and head direction cell supports orientation of the head within the brain.  These cells indicate which direction the head is facing with a frame of reference bound to the features of the proximate environment such as the room the animal is in.  And in the neocortex, they provide orientation details relative to landmarks in the proximate external environment.  Head direction cells maintain our sense of direction as our head and body move.  They are present in the entorhinal cortex, retrosplenial cortex, anterior dorsal and lateral dorsal thalamus, lateral mammillary nucleus, dorsal tegmental nucleus and striatum. 
Plans emerge in complex adaptive systems (CAS) to provide the instructions that agents use to perform actions.  The component architecture and structure of the plans is reviewed. 
schematic strategies
.  Hawkins notes not all of the columns are modeling percepts are internal appearences of the external world and the body according to Haikonen.  RSS views them as evolved models that are:
  • Associated schematically with the signals generated in response to epi-phenomena detected by sensory receptors and
  • Acted on by emergent agents.  
.  These same strategies support all facets of intelligence enables the achievement of goals in the face of obstacles.  The goals are sub-goals of genes' survival and reproduction and include:
  • Obtaining and eating food
  • Sex
  • Finding and maintaining shelter
  • Fighting for resources - in the preferred hunter-gatherer environment loss of resources was critical while possession was often transient. 
  • Understanding the proximate environment
  • Securing the cooperation of others 


69 Concepts, Language, and high-level thinking
Human cognition is the ability to orchestrate thought and action in accordance with internal goals according to Princeton's Jonathan Cohen. 
appears exceptional.  But Mountcastle's proposal suggests all cortical columns must perform very similar functions.  This implies to Hawkins that language and other high-level cognitive is the ability to orchestrate thought and action in accordance with internal goals according to Princeton's Jonathan Cohen. 
abilities are close analogs of seeing, touching and hearing.  He views reference frames is a coordinate system (set of axis) centered on a particular aspect of the situation that describes the location of an object.  It tells you where things are located relative to each other.  And they can detail how to achieve goals such as movements to get from one location to another.  The brain supports many frames of reference, with which, Jeff Hawkins realized, it can model its proximate environment, including for vision: Geons, LIP, superior colliculi; (2009), hearing: gain modulation; & movement planning: PRR; (Jul 2002).  Auditory stimuli are initially coded in a head-centered reference frame.  The motor system codes actions in reference frames that depend on motor effectors.  Eye movements are codes in a reference frame that depends on the difference between current and desired arm position.  It is often necessary to transform the location representation of the sensory stimulus into a representation appropriate for the motor act.  An eye-centered reference frame depends on the location of the eye in the head.  A retinotopic reference frame depends on the retinal location that is activated by a visual stimulus.  Double-saccade tasks show how the location of the second visual target is coded relative to current and desired eye position (eye-centered).  
as ways to organize any kind of knowledge.  Hawkins asserts we treat physical entities are, according to Abbott, a class including people, families, corporations, hurricanes.  They implement abstract designs and are demarcatable by their reduced entropy relative to their components.  Rovelli notes entities are a collection of relations and events, but memory and our continuous process of anticipation, organizes the series of quantized interactions we perceive into an illusion of permanent objects flowing from past to future.  Abbott identifies two types of entity:
  1. At equilibrium entities,
  2. Autonomous entities, which can control how they are affected by outside forces;
we can't touch by picturing them as if we can see them and develop
The agents in complex adaptive systems (CAS) must model their environment to respond effectively to it.  Evolution's schematic operators and Samuel modeling together support the indirect recording of past successes and their strategic use by the current agent to learn how to succeed in the proximate environment. 
models
of them as if we can touch them.  This allows us to use reference frames to organize and think about these items.  Concepts are not three-dimensional objects is a collection of: happenings, occurrences and processes; including emergent entities, as required by relativity, explains Rovelli.  But natural selection has improved our fitness by representing this perception, in our minds, as an unchanging thing, as explained by Pinker.  Dehaene explains the object modeling and construction process within the unconscious and conscious brain.  Hawkins elaborates how the cortical columns of the neocortex identify and locate objects in our proximate environment and generate the stable representation.  Mathematicians view anything that can be defined and used in deductive reasoning and mathematical proofs as an object.  These mathematical objects can be values of variables, allowing them to be used in formulas.  
, but Hawkins notes they can be reasoned about: we predict what will happen to them if
Plans are interpreted and implemented by agents.  This page discusses the properties of agents in a complex adaptive system (CAS). 
It then presents examples of agents in different CAS.  The examples include a computer program where modeling and actions are performed by software agents.  These software agents are aggregates. 
The participation of agents in flows is introduced and some implications of this are outlined. 
we act
in one way or another.  For Hawkins, that implies they must be being managed with reference frames, but ones with more than three dimensions.  He presents a hypothesis:
  1. Reference frames are present everywhere in the neocortex is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 
    - using exaptations, initially termed pre-adaptation refers to the coopting of some function for a new use.  It provides an operational phenotype with schematic access to the adjacent possible through the action of regular genetic operators. 
    of grid, are neurons in the entorhinal cortex, support location modeling.  Grid cells activate when a higher animal reaches a particular position in a triangular grid mapping the floor of the proximate space.  Grid cells are lost early in Alzheimer's disease corresponding to a sense of being lost.  Doeller, Barry and Burgess used fMRI with subjects performing a navigation task: moving around a virtual world on a computer screen; to show that grid like cells also operate in the frontal area of the neocortex.  Constantinescu, O'Reilly and Behrens used fMRI to monitor the frontal neocortex of subjects being shown images of birds with different length necks and legs.  The subjects then imagined new birds that combined the neck of one of the birds with the legs of another.  Grid like cells were observed operating, supporting a reference frame map where one dimension represented neck length and another represented leg length.  And the subjects, in thinking about the birds, were mentally moving through the map.  To think about the birds the neocortex created a new map suited to remembering birds with different necks and legs and "movement" was used to recall the details. 
    , place are neurons which fire whenever a higher animal occupies a certain location in space.  Place cells are highly invariant over a variety of sensory cues, and they even maintain their space-selective firing as the animal wanders around in full darkness.  They encode where the animal thinks it is.  And the same cells associate a time stamp with the memories.   and head direction supports orientation of the head within the brain.  These cells indicate which direction the head is facing with a frame of reference bound to the features of the proximate environment such as the room the animal is in.  And in the neocortex, they provide orientation details relative to landmarks in the proximate external environment.  Head direction cells maintain our sense of direction as our head and body move.  They are present in the entorhinal cortex, retrosplenial cortex, anterior dorsal and lateral dorsal thalamus, lateral mammillary nucleus, dorsal tegmental nucleus and striatum. 
    cell
    Plans emerge in complex adaptive systems (CAS) to provide the instructions that agents use to perform actions.  The component architecture and structure of the plans is reviewed. 
    schematic
    strategies 
  2. Reference frames are used to model everything we know, not just physical objects - cortical columns obtain their context from what provides their inputs.  A column supports a mechanism that blindly tries to discover and model the structure of whatever is causing its inputs to change. 
    This page reviews the implications of selection, variation and heredity in a complex adaptive system (CAS).  The mechanism and its emergence are discussed. 
    Evolution
    leveraged exaptation of schemas for: learning the structure of the environment, to learn about the physical objects we could touch and manipulate, and additionally to learn the structure of concepts
  3. All knowledge is stored at locations relative to reference frames - Every fact we have learned is paired with a location in a reference frame.  Learning details of a subject requires assigning the facts to locations in an appropriate reference frame.  This makes the facts actionable
  4. Thinking is a form of movement - Hawkins sees thinking as neurons, specialized eukaryotic cells include channels which control flows of sodium and potassium ions across the massively extended cell membrane supporting an electro-chemical wave which is then converted into an outgoing chemical signal transmission from synapses which target nearby neuron or muscle cell receptors.  Neurons are supported by glial cells.  Neurons include a:
    • Receptive element - dendrites
    • Transmitting element - axon and synaptic terminals.  The axon may be myelinated, focusing the signals through synaptic transmission, or unmyelinated - where crosstalk is leveraged. 
    • Highly variable DNA schema using transposons. 
    invoking location after location in a reference frame, bringing the mind what was stored at each location.  And this process similarly delivers to the mind the sequence of
    Plans emerge in complex adaptive systems (CAS) to provide the instructions that agents use to perform actions.  The component architecture and structure of the plans is reviewed. 
    strategic actions that achieve some schematic goal

Hawkins notes experimental studies of the brain have identified what and where pathways for each of the senses: vision supports processing of visual data into what and where/how.  To do this it has two distinct paths: The ventral path and the dorsal path. 
, touch, hearing; an example he uses to show how cortical columns can perform different functions by a simple change to their reference frames.  He argues in a 'what' pathway cortical grid cells associate reference frames to projected objects in the environment.  In 'where' pathways the grid like, are neurons in the entorhinal cortex, support location modeling.  Grid cells activate when a higher animal reaches a particular position in a triangular grid mapping the floor of the proximate space.  Grid cells are lost early in Alzheimer's disease corresponding to a sense of being lost.  Doeller, Barry and Burgess used fMRI with subjects performing a navigation task: moving around a virtual world on a computer screen; to show that grid like cells also operate in the frontal area of the neocortex.  Constantinescu, O'Reilly and Behrens used fMRI to monitor the frontal neocortex of subjects being shown images of birds with different length necks and legs.  The subjects then imagined new birds that combined the neck of one of the birds with the legs of another.  Grid like cells were observed operating, supporting a reference frame map where one dimension represented neck length and another represented leg length.  And the subjects, in thinking about the birds, were mentally moving through the map.  To think about the birds the neocortex created a new map suited to remembering birds with different necks and legs and "movement" was used to recall the details. 
cells associate reference frames to our body: looking at a hand, the where visual pathway via the parietal cortex is primarily concerned with space and action.  It attends to distance, position, speed, and orientation in space.  It is not mirror symmetric. 
identifies its location relative to the body, while the what visual pathway focuses on invariant object recognition.  It identifies what is being looked at - leveraging geon based representations of objects.  It is highly sensative to image identity, shape, and color, but ignores size and spatial orientation. 
identifies the model representing what the hand is.  Hawkins suggests the where pathway synapses, a neuron structure which provides a junction with other neurons.  It generates signal molecules, either excitatory or inhibitory, which are kept in vesicles until the synapse is stimulated when the signal molecules are released across the synaptic cleft from the neuron.  The provisioning of synapses is under genetic control and is part of long term memory formation as identified by Eric Kandel.  Modulation signals (from slow receptors) initiate the synaptic strengthening which occurs in memory. 
neurons sensing joint angles in the limbs to the dendrite trees of neurons in its cortical columns while what pathways synapse sense data about the external world and newer internal world to its columns' dendrite trees.  He asserts the where pathways, which are a significant portion of the neocortex, are dedicated to modeling the body and the space around it. 

76 Reference Frames for Concepts
Hawkins stresses that reference frames is a coordinate system (set of axis) centered on a particular aspect of the situation that describes the location of an object.  It tells you where things are located relative to each other.  And they can detail how to achieve goals such as movements to get from one location to another.  The brain supports many frames of reference, with which, Jeff Hawkins realized, it can model its proximate environment, including for vision: Geons, LIP, superior colliculi; (2009), hearing: gain modulation; & movement planning: PRR; (Jul 2002).  Auditory stimuli are initially coded in a head-centered reference frame.  The motor system codes actions in reference frames that depend on motor effectors.  Eye movements are codes in a reference frame that depends on the difference between current and desired arm position.  It is often necessary to transform the location representation of the sensory stimulus into a representation appropriate for the motor act.  An eye-centered reference frame depends on the location of the eye in the head.  A retinotopic reference frame depends on the retinal location that is activated by a visual stimulus.  Double-saccade tasks show how the location of the second visual target is coded relative to current and desired eye position (eye-centered).  
can process concept driven inputs with two changes:
  1. The cortical columns don't have to be fed by a real
    Agents use sensors to detect events in their environment.  This page reviews how these events become signals associated with beneficial responses in a complex adaptive system (CAS).  CAS signals emerge from the Darwinian information model.  Signals can indicate decision summaries and level of uncertainty. 
    sensor
    's outputs
  2. The reference frames dimensionality may be expanded from two or three to four or more; for Hawkins the cortical column's
    The agents in complex adaptive systems (CAS) must model their environment to respond effectively to it.  Evolution's schematic operators and Samuel modeling together support the indirect recording of past successes and their strategic use by the current agent to learn how to succeed in the proximate environment. 
    learning will model
    the best number of dimensions to apply. 
Hawkins explains that becoming a subject area expert: mathematics, politics, language; requires discovering an appropriate reference frame and what movements to perform on the map to reach a goal.  Hawkins concludes different people may organize the concept maps using alternative structures.  Hawkins notes Einstein's examples of trains, people and flashlights to deduce the equations of special relativity, states according to Einstein: if, relative to K, K' is a uniformly moving co-ordinate system devoid of rotation, then natural phenomena run their course with respect to K' according to exactly the same general laws as with respect to K. 
, which allows anyone to follow along.  But general relativity depends on field equations, which Einstein and everyone else struggle to understand. 

Hawkins concludes that Mountcastle's common algorithm is based on reference frames implemented within each of the 150,000 cortical columns. 
83 Language
Wernicke is the posterior part of the left superior temporal cortex.  It is involved in the comprehension of spoken or written language.  Sensations from the ears generate signals to the auditory cortex which then flow on to Wernicke's area.  Wernicke showed the arcuate fasciculus pathway connects this area with Broca's area.  's and Broca is the left inferior frontal cortex.  It is involved with spoken language.  Lesions have resulted in an inability to speak or write even though language is understood.  's areas are associated with language processing, residing on the left side of the brain only, possibly because of the time critical requirements.  Hawkins notes there is disagreement over the location and extent of these two areas and highlights the functionality overlaps.  And he argues large areas of the neocortex is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 
are needed: Wernicke's and Broca's areas don't get direct input from the sensors, the comprehension of language depends on all the auditory and visual regions, and the production of language will depend on different motor networks.  Hawkins stresses that the anatomy of Wernicke's and Broca's areas is similar to other areas of the neocortex.  He assumes the majority of mechanisms underlying language: recursion; are shared with other parts of cognition is the ability to orchestrate thought and action in accordance with internal goals according to Princeton's Jonathan Cohen. 
and
This page discusses the interdependence of perception and representation in a complex adaptive system (CAS).  Hofstadter and Mitchell's research with Copycat is reviewed.  The bridging of a node from a network of 'well known' percepts to a new representational instance is discussed as it occurs in biochemistry, within the brain, in consciousness and abstractly. 
perception
.  Hawkins notes that recursion is used repeatedly in these reference frames is a coordinate system (set of axis) centered on a particular aspect of the situation that describes the location of an object.  It tells you where things are located relative to each other.  And they can detail how to achieve goals such as movements to get from one location to another.  The brain supports many frames of reference, with which, Jeff Hawkins realized, it can model its proximate environment, including for vision: Geons, LIP, superior colliculi; (2009), hearing: gain modulation; & movement planning: PRR; (Jul 2002).  Auditory stimuli are initially coded in a head-centered reference frame.  The motor system codes actions in reference frames that depend on motor effectors.  Eye movements are codes in a reference frame that depends on the difference between current and desired arm position.  It is often necessary to transform the location representation of the sensory stimulus into a representation appropriate for the motor act.  An eye-centered reference frame depends on the location of the eye in the head.  A retinotopic reference frame depends on the retinal location that is activated by a visual stimulus.  Double-saccade tasks show how the location of the second visual target is coded relative to current and desired eye position (eye-centered).  
to represent physical objects is a collection of: happenings, occurrences and processes; including emergent entities, as required by relativity, explains Rovelli.  But natural selection has improved our fitness by representing this perception, in our minds, as an unchanging thing, as explained by Pinker.  Dehaene explains the object modeling and construction process within the unconscious and conscious brain.  Hawkins elaborates how the cortical columns of the neocortex identify and locate objects in our proximate environment and generate the stable representation.  Mathematicians view anything that can be defined and used in deductive reasoning and mathematical proofs as an object.  These mathematical objects can be values of variables, allowing them to be used in formulas.  
.  And it is trivial for the cortical column to do:
Hawkins concludes cortical columns implementation of reference frames is a coordinate system (set of axis) centered on a particular aspect of the situation that describes the location of an object.  It tells you where things are located relative to each other.  And they can detail how to achieve goals such as movements to get from one location to another.  The brain supports many frames of reference, with which, Jeff Hawkins realized, it can model its proximate environment, including for vision: Geons, LIP, superior colliculi; (2009), hearing: gain modulation; & movement planning: PRR; (Jul 2002).  Auditory stimuli are initially coded in a head-centered reference frame.  The motor system codes actions in reference frames that depend on motor effectors.  Eye movements are codes in a reference frame that depends on the difference between current and desired arm position.  It is often necessary to transform the location representation of the sensory stimulus into a representation appropriate for the motor act.  An eye-centered reference frame depends on the location of the eye in the head.  A retinotopic reference frame depends on the retinal location that is activated by a visual stimulus.  Double-saccade tasks show how the location of the second visual target is coded relative to current and desired eye position (eye-centered).  
is the substrate for: learning the structure of the world, where things are, how they move and change, for concrete objects is a collection of: happenings, occurrences and processes; including emergent entities, as required by relativity, explains Rovelli.  But natural selection has improved our fitness by representing this perception, in our minds, as an unchanging thing, as explained by Pinker.  Dehaene explains the object modeling and construction process within the unconscious and conscious brain.  Hawkins elaborates how the cortical columns of the neocortex identify and locate objects in our proximate environment and generate the stable representation.  Mathematicians view anything that can be defined and used in deductive reasoning and mathematical proofs as an object.  These mathematical objects can be values of variables, allowing them to be used in formulas.  
and for abstract concepts. 

91 The Thousand Brains theory of Intelligence
Numenta's goal was to develop a broad theory of the working of the neocortex is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 
.  Initially it focused on a single cortical column revealing:
For Hawkins these details implied the neocortex is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 
must operate differently to the accepted Hierarchy of Features model of the temporal neocortex, which treats vision as a static process, but due to our movement and the eye are major sensors in primates, based on opsins deployed in the retina & especially fovea, signalling the visual system: Superior colliculi, Thalamus (LGN), Primary visual cortex; and indirectly the amygdala.  They also signal [social] emotional state to other people.  And they have implicit censorious power with pictures of eyes encouraging people within their view to act more honorably.  Eyes are poor scanners and use a saccade to present detail slowly to the fovea.  The eye's optical structures and retina are supported by RPE.  Eyes do not connect to the brain through the brain stem and so still operate in locked-in syndrome.  Evo-devo shows eyes have deep homology.  High pressure within the eye can result in glaucoma.  Genetic inheritance can result in retinoblastoma.  Age is associated with AMD.  's saccades are the small jerky steps by which we reorient our eyes gaze. 
, the brain is presented with rapid change in the sensory inputs.  Just like hearing and touch, vision is dynamic and depends on movement.  Hawkins lists inconsistencies and shortcomings of the Hierarchy of Features model:

He called this shift in perspective the Thousand Brain Theory of Intelligence.

95 The New View of the Neocortex
All cortical columns, even in low-level sensory regions, are capable of
The agents in complex adaptive systems (CAS) must model their environment to respond effectively to it.  Evolution's schematic operators and Samuel modeling together support the indirect recording of past successes and their strategic use by the current agent to learn how to succeed in the proximate environment. 
learning
and recognizing complete objects is a collection of: happenings, occurrences and processes; including emergent entities, as required by relativity, explains Rovelli.  But natural selection has improved our fitness by representing this perception, in our minds, as an unchanging thing, as explained by Pinker.  Dehaene explains the object modeling and construction process within the unconscious and conscious brain.  Hawkins elaborates how the cortical columns of the neocortex identify and locate objects in our proximate environment and generate the stable representation.  Mathematicians view anything that can be defined and used in deductive reasoning and mathematical proofs as an object.  These mathematical objects can be values of variables, allowing them to be used in formulas.  
, by integrating its inputs over time
The neocortex is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 
has many complementary
The agents in complex adaptive systems (CAS) must model their environment to respond effectively to it.  Evolution's schematic operators and Samuel modeling together support the indirect recording of past successes and their strategic use by the current agent to learn how to succeed in the proximate environment. 
models
of a particular object, each in a different column.  Individual columns learn hundreds of complex objects, but each column is limited by its inputs: touch, black and white vision, color vision, V1 is primary visual area (V1).  It mainly responds to LGN's signals for thin lines and object contours.  It is located in the occipital lobe.  - detailed, V2 is visual area (V2).  It responds to combinations of lines with curves and inclines from V1's signals.   - larger scale. 

97 Where is knowledge stored in the brain?
All the brain's knowledge is distributed.  Knowledge of an object is a collection of: happenings, occurrences and processes; including emergent entities, as required by relativity, explains Rovelli.  But natural selection has improved our fitness by representing this perception, in our minds, as an unchanging thing, as explained by Pinker.  Dehaene explains the object modeling and construction process within the unconscious and conscious brain.  Hawkins elaborates how the cortical columns of the neocortex identify and locate objects in our proximate environment and generate the stable representation.  Mathematicians view anything that can be defined and used in deductive reasoning and mathematical proofs as an object.  These mathematical objects can be values of variables, allowing them to be used in formulas.  
is distributed across thousands of columns and complementary aggregate
The agents in complex adaptive systems (CAS) must model their environment to respond effectively to it.  Evolution's schematic operators and Samuel modeling together support the indirect recording of past successes and their strategic use by the current agent to learn how to succeed in the proximate environment. 
models
.  Hawkins notes that
This page introduces the complex adaptive system (CAS) theory frame.  The theory provides an organizing framework that is used by 'life.'  It can illuminate and clarify complex situations and be applied flexibly.  It can be used to evaluate and rank models that claim to describe our perceived reality.  It catalogs the laws and strategies which underpin the operation of systems that are based on the interaction of emergent agents.  It highlights the constraints that shape CAS and so predicts their form.  A proposal that does not conform is wrong. 

John Holland's framework for representing complexity is outlined.  Links to other key aspects of CAS theory discussed at the site are presented. 
CAS
work best when knowledge and
Plans are interpreted and implemented by agents.  This page discusses the properties of agents in a complex adaptive system (CAS). 
It then presents examples of agents in different CAS.  The examples include a computer program where modeling and actions are performed by software agents.  These software agents are aggregates. 
The participation of agents in flows is introduced and some implications of this are outlined. 
actions
are distributed among many, but not too many,
Plans are interpreted and implemented by agents.  This page discusses the properties of agents in a complex adaptive system (CAS). 
It then presents examples of agents in different CAS.  The examples include a computer program where modeling and actions are performed by software agents.  These software agents are aggregates. 
The participation of agents in flows is introduced and some implications of this are outlined. 
agents
.  A hierarchic overlay supports control, but must allow some autonomy to achieve robustness.  Each column is a complete sensory-motor system. 

99 The solution to the binding problem
Hawkins asks why we have a singular perception if we have thousands of
The agents in complex adaptive systems (CAS) must model their environment to respond effectively to it.  Evolution's schematic operators and Samuel modeling together support the indirect recording of past successes and their strategic use by the current agent to learn how to succeed in the proximate environment. 
models
? Its because columns vote with the final perception being the consensus reached after voting by the various models agreeing on the core object they partially sense.  Hawkins adds that tactile columns also share their relative position

101 How is voting accomplished in the brain?
Hawkins asserts that long distance axons, a long extension of a neuron which has a membrane constructed to support the uni-directional flow of action potential from the dendritic tree and cell body to the synaptic terminals.   are sent from cells in the column that project what object is a collection of: happenings, occurrences and processes; including emergent entities, as required by relativity, explains Rovelli.  But natural selection has improved our fitness by representing this perception, in our minds, as an unchanging thing, as explained by Pinker.  Dehaene explains the object modeling and construction process within the unconscious and conscious brain.  Hawkins elaborates how the cortical columns of the neocortex identify and locate objects in our proximate environment and generate the stable representation.  Mathematicians view anything that can be defined and used in deductive reasoning and mathematical proofs as an object.  These mathematical objects can be values of variables, allowing them to be used in formulas.  
is being
Agents use sensors to detect events in their environment.  This page reviews how these events become signals associated with beneficial responses in a complex adaptive system (CAS).  CAS signals emerge from the Darwinian information model.  Signals can indicate decision summaries and level of uncertainty. 
sensed
.  Hawkins suggests the column 'broadcasts' what it thinks it is observing.  If it is uncertain, multiple possibilities will be represented.  And the column will receive projections from other columns regarding their guesses.  The most common guesses suppress the others.   In Hawkins's models of voting, columns don't have to signal a vote to every other column: reaching a small number of other randomly selected columns is sufficient, but voting does require a learning phase. 

102 Stability of perception
Column voting stabilizes both touch and visual
This page discusses the interdependence of perception and representation in a complex adaptive system (CAS).  Hofstadter and Mitchell's research with Copycat is reviewed.  The bridging of a node from a network of 'well known' percepts to a new representational instance is discussed as it occurs in biochemistry, within the brain, in consciousness and abstractly. 
perception
.  Having voted the columns agree on what object is a collection of: happenings, occurrences and processes; including emergent entities, as required by relativity, explains Rovelli.  But natural selection has improved our fitness by representing this perception, in our minds, as an unchanging thing, as explained by Pinker.  Dehaene explains the object modeling and construction process within the unconscious and conscious brain.  Hawkins elaborates how the cortical columns of the neocortex identify and locate objects in our proximate environment and generate the stable representation.  Mathematicians view anything that can be defined and used in deductive reasoning and mathematical proofs as an object.  These mathematical objects can be values of variables, allowing them to be used in formulas.  
is being perceived.  The voting neurons, specialized eukaryotic cells include channels which control flows of sodium and potassium ions across the massively extended cell membrane supporting an electro-chemical wave which is then converted into an outgoing chemical signal transmission from synapses which target nearby neuron or muscle cell receptors.  Neurons are supported by glial cells.  Neurons include a:
  • Receptive element - dendrites
  • Transmitting element - axon and synaptic terminals.  The axon may be myelinated, focusing the signals through synaptic transmission, or unmyelinated - where crosstalk is leveraged. 
  • Highly variable DNA schema using transposons. 
in each column provide a stable pattern defining the object and where it is relative to you.  Other neurons in the column change as we move but the voting neurons stay fixed as long as they are sensing the same object.  If it is impossible to resolve the vote, as in the two faces or candlestick illusion, the decision switches periodically between the two candidates. 

104 Attention
Hawkins argues that when we switch is the focusing (orienting) of our mental resources onto a specific piece of information.  Selective attention uses valuations assigned to each potential object of thought by the basal ganglia.  our attention is the mutli-faceted capability allowing access to consciousness.  It includes selective attention, vigilance, allocating attention, goal focus, and meta-awareness.  Hawkins notes its operation involves the thalamus. 
from one aspect of a scene to another, the voting neurons, specialized eukaryotic cells include channels which control flows of sodium and potassium ions across the massively extended cell membrane supporting an electro-chemical wave which is then converted into an outgoing chemical signal transmission from synapses which target nearby neuron or muscle cell receptors.  Neurons are supported by glial cells.  Neurons include a:
  • Receptive element - dendrites
  • Transmitting element - axon and synaptic terminals.  The axon may be myelinated, focusing the signals through synaptic transmission, or unmyelinated - where crosstalk is leveraged. 
  • Highly variable DNA schema using transposons. 
settle on a different object is a collection of: happenings, occurrences and processes; including emergent entities, as required by relativity, explains Rovelli.  But natural selection has improved our fitness by representing this perception, in our minds, as an unchanging thing, as explained by Pinker.  Dehaene explains the object modeling and construction process within the unconscious and conscious brain.  Hawkins elaborates how the cortical columns of the neocortex identify and locate objects in our proximate environment and generate the stable representation.  Mathematicians view anything that can be defined and used in deductive reasoning and mathematical proofs as an object.  These mathematical objects can be values of variables, allowing them to be used in formulas.  
, and we
This page discusses the interdependence of perception and representation in a complex adaptive system (CAS).  Hofstadter and Mitchell's research with Copycat is reviewed.  The bridging of a node from a network of 'well known' percepts to a new representational instance is discussed as it occurs in biochemistry, within the brain, in consciousness and abstractly. 
perceive
both objects are present even though we attend to only one.  He suggests the refocusing of attention involves the thalamus has all the main inputs to the cortex passing through it.  It is massively supplied with return innervations from the cortical regions it routes too.  It does not stand on the route of the main exits from the cortex.  The parafascicular nucleus of the rat thalamus contains relatively high levels of D5 dopamine receptors.  For human vision the primary system connects to the neocortex via, a small part of the thalamus, the LGN.  , but the mechanism is not understood. 
Attention is essential in how the brain learns
The agents in complex adaptive systems (CAS) must model their environment to respond effectively to it.  Evolution's schematic operators and Samuel modeling together support the indirect recording of past successes and their strategic use by the current agent to learn how to succeed in the proximate environment. 
models
.  Each time you shift attention to a different object, the brain determines the object's location relative to the previously attended object, as part of the operation of attention.  These models are often temporary, being updated as we move or the objects are moved.  Hawkins asserts the same learning process is used in the neocortex is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 
for short and long-lasting models. 

106 Hierarchy in the thousand brains theory
Hawkins admits the anatomy of the neocortex is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 
suggests it operates both as a network of voting columns and a hierarchy.  Hawkins proposes the hierarchy is used to assemble objects is a collection of: happenings, occurrences and processes; including emergent entities, as required by relativity, explains Rovelli.  But natural selection has improved our fitness by representing this perception, in our minds, as an unchanging thing, as explained by Pinker.  Dehaene explains the object modeling and construction process within the unconscious and conscious brain.  Hawkins elaborates how the cortical columns of the neocortex identify and locate objects in our proximate environment and generate the stable representation.  Mathematicians view anything that can be defined and used in deductive reasoning and mathematical proofs as an object.  These mathematical objects can be values of variables, allowing them to be used in formulas.  
into hierarchically composed aggregates.  But he admits how this is done is not understood.  Some hierarchic learning may occur within each column.  The answer probably involves understanding how attention is the mutli-faceted capability allowing access to consciousness.  It includes selective attention, vigilance, allocating attention, goal focus, and meta-awareness.  Hawkins notes its operation involves the thalamus. 
works, so Numenta is studying the thalamus has all the main inputs to the cortex passing through it.  It is massively supplied with return innervations from the cortical regions it routes too.  It does not stand on the route of the main exits from the cortex.  The parafascicular nucleus of the rat thalamus contains relatively high levels of D5 dopamine receptors.  For human vision the primary system connects to the neocortex via, a small part of the thalamus, the LGN.  

Hawkins admits much of the brain remains to be understood, but sees the Thousand Brains Theory as an accurate outline on which additional experimental and theoretical details can be coherently added.  In the rest of the book he uses the theory to discuss our future. 

113 Part 2: Machine Intelligence
Hawkins asserts the study of intelligence enables the achievement of goals in the face of obstacles.  The goals are sub-goals of genes' survival and reproduction and include:
  • Obtaining and eating food
  • Sex
  • Finding and maintaining shelter
  • Fighting for resources - in the preferred hunter-gatherer environment loss of resources was critical while possession was often transient. 
  • Understanding the proximate environment
  • Securing the cooperation of others 
and the neocortex is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 
is in a pre-paradigm state.  Hawkins hopes the Thousand Brains Theory will provide that paradigm.  He notes that artificial intelligence has an established paradigm, using common techniques: artificial neural networks; which have supported steady progress in the field.  But Hawkins sees Thousand Brains Theory requiring a revolution in their approach.  But having given a visionary presentation on hand held computing to Intel's engineering community and having been painfully rejected by them, Hawkins typically avoids giving such talks and instead focused for many years on making the future he predicted happen. 

Hawkins sees the Intel situation recurring:
  • He sees the future of AI differently to the leaders of the field
  • He is worried about the direction in which humanity is heading; and his Intel experience suggests he may not be able to change people's beliefs anyway. 
Hawkins has concluded that Artificial Intelligence must adopt principles that more closely mimic the brain, and, like the shift of computers to smartphones,  the shift will be inevitable. 

119 Why there is no "I" in AI
Hawkins notes that since 1956 AI has repeatedly introduced new technologies which promised to lead towards intelligent machines, which induced enthusiasm followed by pessimism as the creations fell short.  demonstrated impressive capabilities.  The current wave of technologies are artificial neural networks used in deep learning is an artificial intelligence approach where engineers deploy data into deep neural networks. 
, which have delivered impressive results.  But most AI researchers don't think today's AI is intelligent enables the achievement of goals in the face of obstacles.  The goals are sub-goals of genes' survival and reproduction and include:
  • Obtaining and eating food
  • Sex
  • Finding and maintaining shelter
  • Fighting for resources - in the preferred hunter-gatherer environment loss of resources was critical while possession was often transient. 
  • Understanding the proximate environment
  • Securing the cooperation of others 
.  He notes:
121 Two paths to AGI
Hawkins explains there have been two paths used by AI researchers to make intelligent machines:
  1. Develop computers that can outperform humans at specific tasks, in the hope that the researchers will discover how to make computers better at every task
  2. Focus on flexibility, rather than aiming to make the tool outperform the human.  It was the strategy used in some of the early waves of AI but proved too difficult.  It proved difficult to represent everyday knowledge, understandable to a five year old, in a useful way for a computer. 
Deep learning networks are representational models that achieve high performance on difficult pattern recognition problems in vision and speech.  But they need specialized training methods such as greedy layerwise pre-training or HF optimization.  Researchers are gaining access to the participation of the individual 'neurons' using: visualization, attribution, dimensionality reduction, interpretability; (Mar 2018)
have no knowledge beyond the statistical matching, where the training data defines the relationships and is associated, by the designer, with labels.  With appropriately developed sensor processing, the training data can be pictures, or alternatively lots of text.  Hawkins believes to achieve the goal of AGI the networks will have to model the world the way the brain does.  But deep learning, while commercially valuable, avoids knowledge representation by using statistics and lots of data.  For Hawkins this is not the path to having the ability of a five-year-old child. 

124 Brains as a model for AI
For Hawkins the brain is the only thing that is intelligent enables the achievement of goals in the face of obstacles.  The goals are sub-goals of genes' survival and reproduction and include:
  • Obtaining and eating food
  • Sex
  • Finding and maintaining shelter
  • Fighting for resources - in the preferred hunter-gatherer environment loss of resources was critical while possession was often transient. 
  • Understanding the proximate environment
  • Securing the cooperation of others 
.  In part he studies brain theory as a necessary step to create intelligent AI.  He asserts the Thousand Brains Theory solves the problem of knowledge representation, because in it the human brain learns a model, building a complex
This page discusses the interdependence of perception and representation in a complex adaptive system (CAS).  Hofstadter and Mitchell's research with Copycat is reviewed.  The bridging of a node from a network of 'well known' percepts to a new representational instance is discussed as it occurs in biochemistry, within the brain, in consciousness and abstractly. 
representation
aligned with what
This page reviews the implications of reproduction initially generating a single initialized child cell.  For multi-cellular organisms this 'cell' must contain all the germ-line schematic structures including for organelles and multi-generational epi-genetic state.  Any microbiome is subsequently integrated during the innovative deployment of this creative event.  Organisms with skeletal infrastructure cannot complete the process of creation of an associated adult mind, until the proximate environment has been sampled during development.  The mechanism and resulting strategic options are discussed. 
we
sense, in interacting with a physical object is a collection of: happenings, occurrences and processes; including emergent entities, as required by relativity, explains Rovelli.  But natural selection has improved our fitness by representing this perception, in our minds, as an unchanging thing, as explained by Pinker.  Dehaene explains the object modeling and construction process within the unconscious and conscious brain.  Hawkins elaborates how the cortical columns of the neocortex identify and locate objects in our proximate environment and generate the stable representation.  Mathematicians view anything that can be defined and used in deductive reasoning and mathematical proofs as an object.  These mathematical objects can be values of variables, allowing them to be used in formulas.  
.  He concludes truly intelligent machines will learn
The agents in complex adaptive systems (CAS) must model their environment to respond effectively to it.  Evolution's schematic operators and Samuel modeling together support the indirect recording of past successes and their strategic use by the current agent to learn how to succeed in the proximate environment. 
models
of the world using map-like reference frames is a coordinate system (set of axis) centered on a particular aspect of the situation that describes the location of an object.  It tells you where things are located relative to each other.  And they can detail how to achieve goals such as movements to get from one location to another.  The brain supports many frames of reference, with which, Jeff Hawkins realized, it can model its proximate environment, including for vision: Geons, LIP, superior colliculi; (2009), hearing: gain modulation; & movement planning: PRR; (Jul 2002).  Auditory stimuli are initially coded in a head-centered reference frame.  The motor system codes actions in reference frames that depend on motor effectors.  Eye movements are codes in a reference frame that depends on the difference between current and desired arm position.  It is often necessary to transform the location representation of the sensory stimulus into a representation appropriate for the motor act.  An eye-centered reference frame depends on the location of the eye in the head.  A retinotopic reference frame depends on the retinal location that is activated by a visual stimulus.  Double-saccade tasks show how the location of the second visual target is coded relative to current and desired eye position (eye-centered).  
analogously to the neocortex is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 


126 Moving from dedicated to universal AI solutions
Initially electronic computers were designed to replace specific tasks of human computers, people doing mathematical calculations.  After Turing showed computations could be performed on a universal machine, a machine specified by mathematician Alan Turing which is the blueprint for the electronic programmable computer.  It consists of an infinite tape on which symbols can be written.  A movable read/write tape head which can move about the tape and write on or read symbols from the tape.  A set of rules that tell the head what to do next. 
, people experimented with a variety of architectures to build computers before market forces settled on Von Neumann, John was a brilliant Hungarian mathematician who published the earliest paper specifying architecture for digital computing.  It ensured this computing architecture was not patentable.  The architecture has a central processing unit (CPU), random access storage addressable by the CPU and a sequencer.  The architecture encourages a serial software architecture that matches the logic of the sequencer and processing operations on program and data.  Von Neumann, his history, computing architecture and some alternative architectures are reviewed by Melanie Mitchell.  's universal Turing machines.  Hawkins predicts the same positive returns, W. Brian Arthur's conception of how high tech products have positive economic feedback as they deploy.  Classical products such as foods have negative returns to scale since they take increasing amounts of land, and distribution infrastructure to support getting them to market.  High tech products typically become easier to produce or gain from platform and network effects of being connected together overcoming the negative effects of scale. 
economic pressure, and need not to lockout unpredictable future requirements - required by robot construction workers sent to Mars where they just have to cope, will force the migration to a general purpose, flexible, AI solution. 

129 When is something intelligent?
Hawkins wonders about the criteria for judging a machine intelligent?  He concludes you have to look inside to know.  Based on the human brain he proposes four criteria at a minimum:
  1. Learning continuously - since our proximate environment changes and we move about.  It generates the
    To benefit from shifts in the environment agents must be flexible.  Being sensitive to environmental signals agents who adjust strategic priorities can constrain their competitors. 
    flexibility
    to leverage learned knowledge and take advantage of changing conditions. 
  2. Learning via movement - allowing us to explore and
    The agents in complex adaptive systems (CAS) must model their environment to respond effectively to it.  Evolution's schematic operators and Samuel modeling together support the indirect recording of past successes and their strategic use by the current agent to learn how to succeed in the proximate environment. 
    model
    the proximate environment. 
  3. Many models - allowing the easy integration of different and new types of moveable
    Agents use sensors to detect events in their environment.  This page reviews how these events become signals associated with beneficial responses in a complex adaptive system (CAS).  CAS signals emerge from the Darwinian information model.  Signals can indicate decision summaries and level of uncertainty. 
    sensor

  4. Using reference frames to store knowledge - enabling thinking, as one 'moves' from one location to another and 'gathers' the associated knowledge

132 Examples of Reference Frames
Some forms of AI have reference frames is a coordinate system (set of axis) centered on a particular aspect of the situation that describes the location of an object.  It tells you where things are located relative to each other.  And they can detail how to achieve goals such as movements to get from one location to another.  The brain supports many frames of reference, with which, Jeff Hawkins realized, it can model its proximate environment, including for vision: Geons, LIP, superior colliculi; (2009), hearing: gain modulation; & movement planning: PRR; (Jul 2002).  Auditory stimuli are initially coded in a head-centered reference frame.  The motor system codes actions in reference frames that depend on motor effectors.  Eye movements are codes in a reference frame that depends on the difference between current and desired arm position.  It is often necessary to transform the location representation of the sensory stimulus into a representation appropriate for the motor act.  An eye-centered reference frame depends on the location of the eye in the head.  A retinotopic reference frame depends on the retinal location that is activated by a visual stimulus.  Double-saccade tasks show how the location of the second visual target is coded relative to current and desired eye position (eye-centered).  
:
Hawkins admits some people question why he did not include consciousness in his criteria

135 When machines are conscious
Hawkins writes most neuroscientists don't talk much about
Consciousness is no longer mysterious.  In this page we use complex adaptive system (CAS) theory to describe the high-level architecture of consciousness, linking sensory networks, low level feelings and genetically conserved and deployed neural structures into a high level scheduler.  Consciousness is evolution's solution to the complex problems of effective, emergent, multi-cellular perception based strategy.  Constrained by emergence and needing to avoid the epistemological problem of starting with a blank slate with every birth, evolution was limited in its options. 

We explain how survival value allows evolution to leverage available tools: sensors, agent relative position, models, perception & representation; to solve the problem of mobile agents responding effectively to their own state and proximate environment.  Evolution did this by providing a genetically constructed framework that can develop into a conscious CAS. 

And we discuss the implications with regard to artificial intelligence, sentient robots, augmented intelligence, and aspects of philosophy. 
consciousness
, assuming that the brain can be studied as a reductionist physical system in a process that will also explain consciousness.  But David Chalmers disagrees, highlighting the hard problem of consciousness
Hawkins does not see why consciousness is beyond explanation.  He proposes to review what brain theory says about some aspects of consciousness: awareness, qualia; and leave the reader to decide what is missing. 

136 Awareness
Hawkins demonstrates that awareness requires we form moment-to-moment memories in the brain includes functionally different types: Declarative, or explicit, (episodic and semantic), Implicit, Procedural, Sequence, Spatial, Temporal, Verbal; Hebb suggested that glutamate receptive neurons learn by (NMDA channel based) synaptic strengthening: short term memory.  This was shown to happen for explicit memory formation in the hippocampus.  This strengthening is sustained by subsequent LTP.  The non-real-time learning and planning processes operate through consciousness using the working memory structures, and then via sleep, the salient ones are consolidated while the rest are destroyed and garbage collected.   of our
Plans are interpreted and implemented by agents.  This page discusses the properties of agents in a complex adaptive system (CAS). 
It then presents examples of agents in different CAS.  The examples include a computer program where modeling and actions are performed by software agents.  These software agents are aggregates. 
The participation of agents in flows is introduced and some implications of this are outlined. 
actions
.  And he notes
Consciousness is no longer mysterious.  In this page we use complex adaptive system (CAS) theory to describe the high-level architecture of consciousness, linking sensory networks, low level feelings and genetically conserved and deployed neural structures into a high level scheduler.  Consciousness is evolution's solution to the complex problems of effective, emergent, multi-cellular perception based strategy.  Constrained by emergence and needing to avoid the epistemological problem of starting with a blank slate with every birth, evolution was limited in its options. 

We explain how survival value allows evolution to leverage available tools: sensors, agent relative position, models, perception & representation; to solve the problem of mobile agents responding effectively to their own state and proximate environment.  Evolution did this by providing a genetically constructed framework that can develop into a conscious CAS. 

And we discuss the implications with regard to artificial intelligence, sentient robots, augmented intelligence, and aspects of philosophy. 
consciousness
requires we form moment-to-moment memories of our thoughts.  If we didn't remember our thoughts we would be unaware of why we are doing something, and the
Plans emerge in complex adaptive systems (CAS) to provide the instructions that agents use to perform actions.  The component architecture and structure of the plans is reviewed. 
plan of actions
.  Accessibility of the past supports our sense of presence and awareness. 

138 Qualia
Hawkins notes our ears, eyes are major sensors in primates, based on opsins deployed in the retina & especially fovea, signalling the visual system: Superior colliculi, Thalamus (LGN), Primary visual cortex; and indirectly the amygdala.  They also signal [social] emotional state to other people.  And they have implicit censorious power with pictures of eyes encouraging people within their view to act more honorably.  Eyes are poor scanners and use a saccade to present detail slowly to the fovea.  The eye's optical structures and retina are supported by RPE.  Eyes do not connect to the brain through the brain stem and so still operate in locked-in syndrome.  Evo-devo shows eyes have deep homology.  High pressure within the eye can result in glaucoma.  Genetic inheritance can result in retinoblastoma.  Age is associated with AMD.  , and skin transmit sense data using neurons, specialized eukaryotic cells include channels which control flows of sodium and potassium ions across the massively extended cell membrane supporting an electro-chemical wave which is then converted into an outgoing chemical signal transmission from synapses which target nearby neuron or muscle cell receptors.  Neurons are supported by glial cells.  Neurons include a:
  • Receptive element - dendrites
  • Transmitting element - axon and synaptic terminals.  The axon may be myelinated, focusing the signals through synaptic transmission, or unmyelinated - where crosstalk is leveraged. 
  • Highly variable DNA schema using transposons. 
and action potentials are the actively generated waves of voltage change across the neuron's membrane that flow down a neuron's axon.  Helmholtz noted that while they propagate far more slowly than electrical transmissions action potentials do not attenuate.  Lord Adrian showed the action potential to be an all-or-nothing signal.  Consequently Adrian extended the neuron doctrine from anatomy to function demonstrating:
  • Sensitivity is indicated by the frequency of transmission of action potentials. 
  • Anatomy indicates the meaning of the signal.  Hodgkin and Huxley's ionic hypothesis completed the characterization of the action potential.  
that look the same.  But vision, and hearing feel are subjective models: sad, glad, mad, scared, surprised, and compassionate; of the organism and its proximate environment, including ratings of situations signalled by broadly distributed chemicals and neural circuits.  These feelings become highly salient inputs, evolutionarily associated, to higher level emotions encoded in neural circuits: amygdala, and insula.  Deacon shows James' conception of feeling can build sentience.  Damasio, similarly, asserts feelings reveal to the conscious mind the subjective status of life: good, bad, in between; within a higher organism.  They especially indicate the affective situation within the old interior world of the viscera located in the abdomen, thorax and thick of the skin - so smiling makes one feel happy; but augmented with the reports from the situation of the new interior world of voluntary muscles.  Repeated experiences build intermediate narratives, in the mind, which reduce the salience.  Damasio concludes feelings relate closely and consistently with homeostasis, acting as its mental deputies once organisms developed 'nervous systems' about 600 million years ago, and building on the precursor regulatory devices supplied by evolution to social insects and prokaryotes and leveraging analogous dynamic constraints.  Damasio suggests feelings contribute to the development of culture:
  • As motives for intellectual creation: prompting detection and diagnosis of homeostatic deficiencies, identifying desirable states worthy of creative effort.
  • As monitors of the success and failure of cultural instruments and practices
  • As participants in the negotiation of adjustments required by the cultural process over time 
different, and are not
This page discusses the interdependence of perception and representation in a complex adaptive system (CAS).  Hofstadter and Mitchell's research with Copycat is reviewed.  The bridging of a node from a network of 'well known' percepts to a new representational instance is discussed as it occurs in biochemistry, within the brain, in consciousness and abstractly. 
perceived
as arriving spikes.  Instead the sense data is perceived as qualia are the direct qualities of percepts according to Haikonen.  He argues they do not require interpretation or any evocation of meaning.  Colors are colors and pain is pain.  The human visual hierarchy seems at odds with this interpretation with meaning being associated with letters by signalling from the letterbox to the frontal lobes and used in the feedback flows that identify and prime morphemes.  Damasio suggests qualia are a type of provoked feeling, triggered by stimuli like taste or vision, which results in an emotive response. 
, arguably one of the mysteries of
Consciousness is no longer mysterious.  In this page we use complex adaptive system (CAS) theory to describe the high-level architecture of consciousness, linking sensory networks, low level feelings and genetically conserved and deployed neural structures into a high level scheduler.  Consciousness is evolution's solution to the complex problems of effective, emergent, multi-cellular perception based strategy.  Constrained by emergence and needing to avoid the epistemological problem of starting with a blank slate with every birth, evolution was limited in its options. 

We explain how survival value allows evolution to leverage available tools: sensors, agent relative position, models, perception & representation; to solve the problem of mobile agents responding effectively to their own state and proximate environment.  Evolution did this by providing a genetically constructed framework that can develop into a conscious CAS. 

And we discuss the implications with regard to artificial intelligence, sentient robots, augmented intelligence, and aspects of philosophy. 
consciousness


139 Qualia are part of the brain's model of the world
Qualia are the direct qualities of percepts according to Haikonen.  He argues they do not require interpretation or any evocation of meaning.  Colors are colors and pain is pain.  The human visual hierarchy seems at odds with this interpretation with meaning being associated with letters by signalling from the letterbox to the frontal lobes and used in the feedback flows that identify and prime morphemes.  Damasio suggests qualia are a type of provoked feeling, triggered by stimuli like taste or vision, which results in an emotive response. 
are subjective internal experiences, which abstract away the variable effects of our proximate environment.  Hawkins argues that some qualia are learned by movement with
The agents in complex adaptive systems (CAS) must model their environment to respond effectively to it.  Evolution's schematic operators and Samuel modeling together support the indirect recording of past successes and their strategic use by the current agent to learn how to succeed in the proximate environment. 
models
being used to predict what should be sensed as the object shifts, and then with a match become associated via an orientation reference frame is a coordinate system (set of axis) centered on a particular aspect of the situation that describes the location of an object.  It tells you where things are located relative to each other.  And they can detail how to achieve goals such as movements to get from one location to another.  The brain supports many frames of reference, with which, Jeff Hawkins realized, it can model its proximate environment, including for vision: Geons, LIP, superior colliculi; (2009), hearing: gain modulation; & movement planning: PRR; (Jul 2002).  Auditory stimuli are initially coded in a head-centered reference frame.  The motor system codes actions in reference frames that depend on motor effectors.  Eye movements are codes in a reference frame that depends on the difference between current and desired arm position.  It is often necessary to transform the location representation of the sensory stimulus into a representation appropriate for the motor act.  An eye-centered reference frame depends on the location of the eye in the head.  A retinotopic reference frame depends on the retinal location that is activated by a visual stimulus.  Double-saccade tasks show how the location of the second visual target is coded relative to current and desired eye position (eye-centered).  
rather than a location.  Hawkins writes that other qualia: the feeling are subjective models: sad, glad, mad, scared, surprised, and compassionate; of the organism and its proximate environment, including ratings of situations signalled by broadly distributed chemicals and neural circuits.  These feelings become highly salient inputs, evolutionarily associated, to higher level emotions encoded in neural circuits: amygdala, and insula.  Deacon shows James' conception of feeling can build sentience.  Damasio, similarly, asserts feelings reveal to the conscious mind the subjective status of life: good, bad, in between; within a higher organism.  They especially indicate the affective situation within the old interior world of the viscera located in the abdomen, thorax and thick of the skin - so smiling makes one feel happy; but augmented with the reports from the situation of the new interior world of voluntary muscles.  Repeated experiences build intermediate narratives, in the mind, which reduce the salience.  Damasio concludes feelings relate closely and consistently with homeostasis, acting as its mental deputies once organisms developed 'nervous systems' about 600 million years ago, and building on the precursor regulatory devices supplied by evolution to social insects and prokaryotes and leveraging analogous dynamic constraints.  Damasio suggests feelings contribute to the development of culture:
  • As motives for intellectual creation: prompting detection and diagnosis of homeostatic deficiencies, identifying desirable states worthy of creative effort.
  • As monitors of the success and failure of cultural instruments and practices
  • As participants in the negotiation of adjustments required by the cultural process over time 
of pain emerged as a mental experience, Damasio asserts, constructed by the mind using mapping structures and events provided by nervous systems.  But feeling pain is supported by older biological functions that support homeostasis.  These capabilities reflect the organism's underlying emotive processes that respond to wounds: antibacterial and analgesic chemical deployment, flinching and evading actions; that occur in organisms without nervous systems.  Later in evolution, after organisms with nervous systems were able to map non-neural events, the components of this complex response were 'imageable'.  Today, a wound induced by an internal disease is reported by old, unmyelinated C nerve fibers.  A wound created by an external cut is signalled by evolutionarily recent myelinated fibers that result in a sharp well-localized report, that initially flows to the dorsal root ganglia, then to the spinal cord, where the signals are mixed within the dorsal and ventral horns, and then are transmitted to the brain stem nuclei, thalamus and cerebral cortex.  The pain of a cut is located, but it is also felt through an emotive response that stops us in our tracks.  Pain amplifies the aggression response of people by interoceptive signalling of brain regions providing social emotions including the PAG projecting to the amygdala; making aggressive people more so and less aggressive people less so.  Fear of pain is a significant contributor to female anxiety.  Pain is the main reason people visit the ED in the US.  Pain is mediated by the thalamus and nucleus accumbens, unless undermined by sleep deprivation. 
; are not learned in the neocortex is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 


141 The neuroscience of consciousness
While some neuroscientists assume that
Consciousness is no longer mysterious.  In this page we use complex adaptive system (CAS) theory to describe the high-level architecture of consciousness, linking sensory networks, low level feelings and genetically conserved and deployed neural structures into a high level scheduler.  Consciousness is evolution's solution to the complex problems of effective, emergent, multi-cellular perception based strategy.  Constrained by emergence and needing to avoid the epistemological problem of starting with a blank slate with every birth, evolution was limited in its options. 

We explain how survival value allows evolution to leverage available tools: sensors, agent relative position, models, perception & representation; to solve the problem of mobile agents responding effectively to their own state and proximate environment.  Evolution did this by providing a genetically constructed framework that can develop into a conscious CAS. 

And we discuss the implications with regard to artificial intelligence, sentient robots, augmented intelligence, and aspects of philosophy. 
consciousness
is beyond scientific explanation, and some study correlates instead, Hawkins believes science can explain it.  He likes
In Gray Matter Michael Graziano asks Are we Really Conscious?  He argues that we build inaccurate models of reality and then develop intuitions based on these problematic models.  He concludes we can't use intuitions to understand consciousness.  Instead he promotes 'brain science' as more accurate and argues it suggests we are not conscious.  In this page we summarize his article and then use complex adaptive system (CAS) theory to review his arguments.  Constrained by CAS theory and mechanisms of emergence we see a requirement for consciousness. 
Graziano's approach
which is based on the neocortex is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 
The agents in complex adaptive systems (CAS) must model their environment to respond effectively to it.  Evolution's schematic operators and Samuel modeling together support the indirect recording of past successes and their strategic use by the current agent to learn how to succeed in the proximate environment. 
learning a model
of attention is the mutli-faceted capability allowing access to consciousness.  It includes selective attention, vigilance, allocating attention, goal focus, and meta-awareness.  Hawkins notes its operation involves the thalamus. 


142 Machine consciousness
Hawkins argues, if
Consciousness is no longer mysterious.  In this page we use complex adaptive system (CAS) theory to describe the high-level architecture of consciousness, linking sensory networks, low level feelings and genetically conserved and deployed neural structures into a high level scheduler.  Consciousness is evolution's solution to the complex problems of effective, emergent, multi-cellular perception based strategy.  Constrained by emergence and needing to avoid the epistemological problem of starting with a blank slate with every birth, evolution was limited in its options. 

We explain how survival value allows evolution to leverage available tools: sensors, agent relative position, models, perception & representation; to solve the problem of mobile agents responding effectively to their own state and proximate environment.  Evolution did this by providing a genetically constructed framework that can develop into a conscious CAS. 

And we discuss the implications with regard to artificial intelligence, sentient robots, augmented intelligence, and aspects of philosophy. 
consciousness
is just a physical phenomena then, machines that work the same way as the brain will be conscious.  As are many animals: mammals.  And he sees turning off a conscious machine as analogous to its going to sleep facilitates salient memory formation and removal of non-salient memories.  The five different stages of the nightly sleep cycles support different aspects of memory formation.  The sleep stages follow Pre-sleep and include: Stage one characterized by light sleep and lasting 10 minutes, Stage two where theta waves and sleep spindles occur, Stage three and Stage four together represent deep slow-wave sleep (SWS) with delta waves, Stage five is REM sleep; sleep cycles last between 90-110 minutes each and as the night progresses SWS times reduce and REM times increase.   Sleep includes the operation of synapse synthesis and maintenance through DNA based activity including membrane trafficking, synaptic vesicle recycling, myelin structural protein formation and cholesterol and protein synthesis.  Sleep also controls inflammation (Jan 2019)  Sleep deprivation undermines the thalamus & nucleus accumbens management of pain. 
rather than murder.  He does not even consider its destruction as murder.  Hawkins sees fear is an emotion which prepares the body for time sensitive action: Blood is sent to the muscles from the gut and skin, Adrenalin is released stimulating: Fuel to be released from the liver, Blood is encouraged to clot, and Face is wide-eyed and fearful.  The short-term high priority goal, experienced as a sense of urgency, is to flee, fight or deflect the danger.  There are both 'innate' - really high priority learning - which are mediated by the central amygdala and learned fears which are mediated by the BLA which learns to fear a stimulus and then signals the central amygdala.  Tara Brach notes we experience fear as a painfully constricted throat, chest and belly, and racing heart.  The mind can build stories of the future which include fearful situations making us anxious about current ideas and actions that we associate with the potential future scenario.  And it can associate traumatic events from early childhood with our being at fault.  Consequent assumptions of our being unworthy can result in shame and fear of losing friendships.  The mechanism for human fear was significantly evolved to protect us in the African savanna.  This does not align perfectly with our needs in current environments: U.S. Grant was unusually un-afraid of the noise or risk of guns and trusted his horses' judgment, which mostly benefited his agency as a modern soldier. 
of death driven by the old brain's reflex emotional are low level fast unconscious agents distributed across the brain and body which associate, via the amygdala and rich club hubs, important environmental signals with encoded high speed sensors, and distributed programs of action to model: predict, prioritize guidance signals, select and respond effectively, coherently and rapidly to the initial signal.  The majority of emotion centered brain regions interface to the midbrain through the hypothalamus.  The cerebellum and basal ganglia support the integration of emotion and motor functions, rewarding rhythmic movement.  The most accessible signs of emotions are the hard to control and universal facial expressions.  Emotions provide prioritization for conscious access given that an animal has only one body, but possibly many cells, with which to achieve its highest level goals.  Because of this, base emotions clash with group goals and are disparaged by the powerful.  Pinker notes a set of group selected emotions which he classes as: other-condemning, other-praising, other-suffering and self-conscious emotions.  Evolutionary psychology argues evolution shaped human emotions during the long period of hunter-gatherer existence in the African savanna.  Human emotions are universal and include: Anger, Appreciation of natural beauty, Contempt, Disgust, Embarrassment, Fear, Gratitude, Grief, Guilt, Happiness, Honor, Jealousy, Liking, Love, Moral awe, Rage, Romantic love, Lust for revenge, Passion, Sadness, Self-control, Shame, Sympathy, Surprise; and the sham emotions and distrust induced by reciprocal altruism. 
response is according to Damasio, a process including a collection of actions: release of specific chemicals in sites of the CNS or their transport, by neural signalling to varied regions of the nervous system and body.  Endocrine glands are signalled and produce molecules capable of altering body function; altering viscera, that changes the homeostatic state of the organism, and may change the spontaneous feelings too.  A cascade of spontaneous homeostatic changes: metabolism, nervous system, immune response, mind builds 'images'; becomes an ensemble of actions each represented in the mind, summarized as a provoked feeling.  Attention to the feelings varies depending on the current state of the mind.  Emotive responses are generated non consciously by specific nuclei in the brain:
  • Hypothalamic nuclei
  • PAG
  • Amygdala nuclei and nucleus accumbens; each nuclei activated by particular streams of signals, from the senses or memory, enabling responses to vast numbers of sensations, objects and circumstances with drives, motivations and emotions. 
.  He notes without the old brain we would not sense fear or sadness is a feeling, which can induce empathy and compassion.  It can last for days, in contrast to the emotions, fear & anger.  Mild sadness induces a beneficial state in the brain: improved judgment, memory, motivation, and more socially sensitive and generous. 
.  And he asserts fear of death and sorrow for loss, are not required ingredients for a machine to be conscious or intelligent

143 The mystery of life and the mystery of consciousness
What is life? was until recently a deep mystery and it was presumed could not be understood scientifically.  It is now clear, that was a misconception.  Hawkins sees the same shift occurring for
Consciousness is no longer mysterious.  In this page we use complex adaptive system (CAS) theory to describe the high-level architecture of consciousness, linking sensory networks, low level feelings and genetically conserved and deployed neural structures into a high level scheduler.  Consciousness is evolution's solution to the complex problems of effective, emergent, multi-cellular perception based strategy.  Constrained by emergence and needing to avoid the epistemological problem of starting with a blank slate with every birth, evolution was limited in its options. 

We explain how survival value allows evolution to leverage available tools: sensors, agent relative position, models, perception & representation; to solve the problem of mobile agents responding effectively to their own state and proximate environment.  Evolution did this by providing a genetically constructed framework that can develop into a conscious CAS. 

And we discuss the implications with regard to artificial intelligence, sentient robots, augmented intelligence, and aspects of philosophy. 
consciousness
, when the hard problem will not exist. 

145 The future of machine intelligence
There are no technical problems preventing the creation of intelligent enables the achievement of goals in the face of obstacles.  The goals are sub-goals of genes' survival and reproduction and include:
  • Obtaining and eating food
  • Sex
  • Finding and maintaining shelter
  • Fighting for resources - in the preferred hunter-gatherer environment loss of resources was critical while possession was often transient. 
  • Understanding the proximate environment
  • Securing the cooperation of others 
machines.  The obstacle was lack of understanding of what intelligence is.  Hawkins is convinced the deployment of intelligent machines will have a transformative impact.  How that will occur is uncertain is when a factor is hard to measure because it is dependent on many interconnected agents and may be affected by infrastructure and evolved amplifiers.  This is different from risk, although the two are deliberately conflated by ERISA.  Keynes argued that most aspects of the future are uncertain, at best represented by ordinal probabilities, and often only by capricious hope for future innovation, fear inducing expectations of limited confidence, which evolutionary psychology implies is based on the demands of our hunter gatherer past.  Deacon notes reduced uncertainty equates to information. 
.  But he feels the Thousand Brains Theory will help define the boundaries of what is possible:

161 The existential risks of machine intelligence
Hawkins notes how quickly opinion shifted from "AI might never happen" to "AI is likely to destroy all humans in the near future."  And it is an opinion expressed by leading technologists, scientists and philosophers encouraging many to fear is an emotion which prepares the body for time sensitive action: Blood is sent to the muscles from the gut and skin, Adrenalin is released stimulating: Fuel to be released from the liver, Blood is encouraged to clot, and Face is wide-eyed and fearful.  The short-term high priority goal, experienced as a sense of urgency, is to flee, fight or deflect the danger.  There are both 'innate' - really high priority learning - which are mediated by the central amygdala and learned fears which are mediated by the BLA which learns to fear a stimulus and then signals the central amygdala.  Tara Brach notes we experience fear as a painfully constricted throat, chest and belly, and racing heart.  The mind can build stories of the future which include fearful situations making us anxious about current ideas and actions that we associate with the potential future scenario.  And it can associate traumatic events from early childhood with our being at fault.  Consequent assumptions of our being unworthy can result in shame and fear of losing friendships.  The mechanism for human fear was significantly evolved to protect us in the African savanna.  This does not align perfectly with our needs in current environments: U.S. Grant was unusually un-afraid of the noise or risk of guns and trusted his horses' judgment, which mostly benefited his agency as a modern soldier. 
AI is an existential threat, due to:

All these scenarios depend on a misunderstanding of: intelligence enables the achievement of goals in the face of obstacles.  The goals are sub-goals of genes' survival and reproduction and include:
  • Obtaining and eating food
  • Sex
  • Finding and maintaining shelter
  • Fighting for resources - in the preferred hunter-gatherer environment loss of resources was critical while possession was often transient. 
  • Understanding the proximate environment
  • Securing the cooperation of others 
, and what is technically feasible.  Intelligence explosions are highly constrained by the need to learn and the opportunity cost and time knowledge acquisition takes.  Similarly superhuman intelligence is constrained: by the way new theories becomes accepted.  It takes years to move from the initial group, to being accepted by other initially critical peers, until years later it becomes generally accepted, and because the proximate environment is continuously changing and no one can be everywhere at once.  Current deep learning is an artificial intelligence approach where engineers deploy data into deep neural networks. 
scenarios take advantage of situations that are stable.  Goal-misalignment requires two improbable events: the intelligent machine can only accept one goal, and it is capable of success in achieving the goal even as we try to stop it. 

And Hawkins's view that intelligent machines don't pose an existential threat has been criticized by those who see the risk, is an assessment of the likelihood of an independent problem occurring.  It can be assigned an accurate probability since it is independent of other variables in the system.  As such it is different from uncertainty. 
as analogous to the power asymmetry that led to the destruction of indigenous tribes by colonizing Europeans.  But Hawkins's proposed design of intelligent machines old brain equivalent specifically excludes the psychological basis of senses and emotions from apes: self-replication, and its supporting
This page reviews the implications of selection, variation and heredity in a complex adaptive system (CAS).  The mechanism and its emergence are discussed. 
evolved
This page reviews the strategy of setting up an arms race.  At its core this strategy depends on being able to alter, or take advantage of an alteration in, the genome or equivalent.  The situation is illustrated with examples from biology, high tech and politics. 
motivations
; that has proven
E. O. Wilson reviews the effect of man on the natural world to date and explains how the two systems can coexist most effectively. 
increasingly problematic
, while he sees intelligence as relatively benign. 

Hawkins agrees there are scenarios where intelligent machines will enhance problematic human strategies.  But does not consider these to be existential threats. 

Hawkins sees machine intelligence as a powerful technology that can be helpful in humanity's future.  He does not see instances of his architecture acting independently as an existential threat. 
171 Part 3: Human intelligence
Hawkins notes the
E. O. Wilson reviews the effect of man on the natural world to date and explains how the two systems can coexist most effectively. 
massive destructive changes
in the operations of the planet and the biosphere attributable to human intelligence.  Our intelligence has allowed us to
Matt Ridley demonstrates the creative effect of man on the World. He highlights:
  • A list of preconditions resulting in
  • Additional niche capture & more free time 
  • Building a network to interconnect memes processes & tools which
  • Enabling inter-generational transfers
  • Innovations that help reduce environmental stress even as they leverage fossil fuels

achieve remarkable success based on leverage of technology and scientific discoveries
.  Our very success has contributed to the scale of problems we are causing: eight billion people doubling every 20 years, polluting every part of the planet, and nuclear weapons and gene editing have the potential to allow a few people to kill billions.  Hawkins concludes our nature has become an existential threat.  He, wants to start a conversation about issues and opportunities made available by leveraging the Thousand Brains Theory.  To do so he reviews: the inherent risks, is an assessment of the likelihood of an independent problem occurring.  It can be assigned an accurate probability since it is independent of other variables in the system.  As such it is different from uncertainty. 
generated by the way the human brain is structured: simulation allows false beliefs, our old brain's nature and the neocortex is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 
's mistakes; existing strategies to mitigate the risks framed by brain theory, and new ideas he thinks might help. 

173 False beliefs
The world we experience is not the physical one in which we live, but instead is a simulation, constructed by our brains to make us effective in coping with our proximate environment.  But we are often misled.  The brain is a
This page discusses the effect of the network on the agents participating in a complex adaptive system (CAS).  Small world and scale free networks are considered. 
network
of neurons, specialized eukaryotic cells include channels which control flows of sodium and potassium ions across the massively extended cell membrane supporting an electro-chemical wave which is then converted into an outgoing chemical signal transmission from synapses which target nearby neuron or muscle cell receptors.  Neurons are supported by glial cells.  Neurons include a:
  • Receptive element - dendrites
  • Transmitting element - axon and synaptic terminals.  The axon may be myelinated, focusing the signals through synaptic transmission, or unmyelinated - where crosstalk is leveraged. 
  • Highly variable DNA schema using transposons. 
connected to
Agents use sensors to detect events in their environment.  This page reviews how these events become signals associated with beneficial responses in a complex adaptive system (CAS).  CAS signals emerge from the Darwinian information model.  Signals can indicate decision summaries and level of uncertainty. 
sensors
and muscles in the body.  The qualia are the direct qualities of percepts according to Haikonen.  He argues they do not require interpretation or any evocation of meaning.  Colors are colors and pain is pain.  The human visual hierarchy seems at odds with this interpretation with meaning being associated with letters by signalling from the letterbox to the frontal lobes and used in the feedback flows that identify and prime morphemes.  Damasio suggests qualia are a type of provoked feeling, triggered by stimuli like taste or vision, which results in an emotive response. 
that we subjectively feel are subjective models: sad, glad, mad, scared, surprised, and compassionate; of the organism and its proximate environment, including ratings of situations signalled by broadly distributed chemicals and neural circuits.  These feelings become highly salient inputs, evolutionarily associated, to higher level emotions encoded in neural circuits: amygdala, and insula.  Deacon shows James' conception of feeling can build sentience.  Damasio, similarly, asserts feelings reveal to the conscious mind the subjective status of life: good, bad, in between; within a higher organism.  They especially indicate the affective situation within the old interior world of the viscera located in the abdomen, thorax and thick of the skin - so smiling makes one feel happy; but augmented with the reports from the situation of the new interior world of voluntary muscles.  Repeated experiences build intermediate narratives, in the mind, which reduce the salience.  Damasio concludes feelings relate closely and consistently with homeostasis, acting as its mental deputies once organisms developed 'nervous systems' about 600 million years ago, and building on the precursor regulatory devices supplied by evolution to social insects and prokaryotes and leveraging analogous dynamic constraints.  Damasio suggests feelings contribute to the development of culture:
  • As motives for intellectual creation: prompting detection and diagnosis of homeostatic deficiencies, identifying desirable states worthy of creative effort.
  • As monitors of the success and failure of cultural instruments and practices
  • As participants in the negotiation of adjustments required by the cultural process over time 
, can only exist in the brain's
The agents in complex adaptive systems (CAS) must model their environment to respond effectively to it.  Evolution's schematic operators and Samuel modeling together support the indirect recording of past successes and their strategic use by the current agent to learn how to succeed in the proximate environment. 
model
of the world, which only represents phenomena that are detected by our specific set of sensors.  Hawkins stresses:
  1. The brain only knows about a subset of the real world
  2. We perceive our model of the world, not the world itself; Hawkins aims to explain how this leads to false beliefs, and what can be done about this. 
The model can be wrong.  We can perceive things that don't exist, and incorrectly perceive things that do exist.  Most of the time this is useful because the brain's model has evolved strategic assumptions about the world to cope with ill posed problems presented to our senses.  But sometimes the model is wrong: beliefs are evolved to help us be fit rather than being shaped for truth

Plans emerge in complex adaptive systems (CAS) to provide the instructions that agents use to perform actions.  The component architecture and structure of the plans is reviewed. 
Memes
are designed to leverage our brains modeling system, making them easy to
The agents in complex adaptive systems (CAS) must model their environment to respond effectively to it.  Evolution's schematic operators and Samuel modeling together support the indirect recording of past successes and their strategic use by the current agent to learn how to succeed in the proximate environment. 
learn
, and transmit to others. 

Incorrect meme-structured models are particularly troublesome, Hawkins concludes.  He notes if they help their believers to be fit is, according to Dawkins, a suitcase word with at least five meanings in biology:
  1. Darwin and Wallace thought in terms of the capacity to survive and reproduce, but they were considering discrete aspects such as chewing grass - where hard enamel would improve the relative fitness. 
  2. Population geneticists: Ronald Fisher, Sewall Wright, J.B.S. Haldane; consider selection at a locus where for a genotype: green eyes vs blue eyes; one with higher fitness can be identified from genotypic frequencies and gene frequencies, with all other variations averaged out. 
  3. Whole organism 'integrated' fitness.  Dawkins notes there is only ever one instance of a specific organism.  Being unique, comparing the relative success of its offspring makes little sense.  Over a huge number of generations the individual is likely to have provided a contribution to everyone in the pool or no one. 
  4. Inclusive fitness, where according to Hamilton, fitness depends on an organism's actions or effects on its children or its relative's children, a model where natural selection favors organs and behaviors that cause the individual's genes to be passed on.  It is easy to mistakenly count an offspring in multiple relative's fitness assessments. 
  5. Personal fitness represents the effects a person's relatives have on the individual's fitness [3].  When interpreted correctly fitness [4] and fitness [5] are the same. 
, the meme will also succeed, in a mutually reinforcing, co-evolution.  Both succeed with increased replication. 

Hawkins asserts that language is an amplifier of false beliefs, since it extends people's models of the world without their having to personally experience new situations.  Based on trust and distrust are evolved responses to sham emotions.  During a friendship where no sham emotions have been detected trust will build up. 
, a massive expansion of knowledge became possible.  But
The agents in complex adaptive systems (CAS) must model their environment to respond effectively to it.  Evolution's schematic operators and Samuel modeling together support the indirect recording of past successes and their strategic use by the current agent to learn how to succeed in the proximate environment. 
learning
Rather than oppose the direct thrust of some environmental flow agents can improve their effectiveness with indirect responses.  This page explains how agents are architected to do this and discusses some examples of how it can be done. 
indirectly
with language is not totally reliable.  And to validate what we have learned requires costly searching for contrary evidence.  Until some contrary detail is found, we have to assume the
The agents in complex adaptive systems (CAS) must model their environment to respond effectively to it.  Evolution's schematic operators and Samuel modeling together support the indirect recording of past successes and their strategic use by the current agent to learn how to succeed in the proximate environment. 
model
is legitimate.  But Hawkins notes many people do not abandon models from their in-group that have been shown to be false, and the neocortex is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 
can self-correct but is thwarted, he argues, by viral false beliefs

185 The existential risks of human intelligence
Hawkins asserts intelligence enables the achievement of goals in the face of obstacles.  The goals are sub-goals of genes' survival and reproduction and include:
  • Obtaining and eating food
  • Sex
  • Finding and maintaining shelter
  • Fighting for resources - in the preferred hunter-gatherer environment loss of resources was critical while possession was often transient. 
  • Understanding the proximate environment
  • Securing the cooperation of others 
itself is benign, once the selfish drives, motivations, and emotions are low level fast unconscious agents distributed across the brain and body which associate, via the amygdala and rich club hubs, important environmental signals with encoded high speed sensors, and distributed programs of action to model: predict, prioritize guidance signals, select and respond effectively, coherently and rapidly to the initial signal.  The majority of emotion centered brain regions interface to the midbrain through the hypothalamus.  The cerebellum and basal ganglia support the integration of emotion and motor functions, rewarding rhythmic movement.  The most accessible signs of emotions are the hard to control and universal facial expressions.  Emotions provide prioritization for conscious access given that an animal has only one body, but possibly many cells, with which to achieve its highest level goals.  Because of this, base emotions clash with group goals and are disparaged by the powerful.  Pinker notes a set of group selected emotions which he classes as: other-condemning, other-praising, other-suffering and self-conscious emotions.  Evolutionary psychology argues evolution shaped human emotions during the long period of hunter-gatherer existence in the African savanna.  Human emotions are universal and include: Anger, Appreciation of natural beauty, Contempt, Disgust, Embarrassment, Fear, Gratitude, Grief, Guilt, Happiness, Honor, Jealousy, Liking, Love, Moral awe, Rage, Romantic love, Lust for revenge, Passion, Sadness, Self-control, Shame, Sympathy, Surprise; and the sham emotions and distrust induced by reciprocal altruism. 
of human intelligence are excluded.  Actual human intelligence includes these aspects and over the last hundred years has become a threat, being capable of initiating integrates a:
  • Signal that advertises the presence of the doomsday machine
  • Machine that once started can't be stopped.  
  • Uncontrollable initiation of the machine based on some constraint. 
  • Catastrophic result for all parties once the machine is started.  There is the potential for both parties to participate in an arms race. 
nuclear war, and inducing climate change based extinction events.  Inducing more such threats seem possible unless we systemically remove the cause or compensate for the threat.  Hawkins highlights two systemic risks, is an assessment of the likelihood of an independent problem occurring.  It can be assigned an accurate probability since it is independent of other variables in the system.  As such it is different from uncertainty. 
associated with the brain:
  1. The old brain evolved to support our primitive desires and actions and is often selfish and shortsighted - brains supported the gene's need to survive until replicated.  From the perspective of individuals and
    The specialized environment and evolution of humans on the African savanna supports the development of a new type of superOrganism.  The emergence of culture allowed human superOrganism families to accelerate the evolutionary process and apply it to memes.  This cultural superOrganism can evolve significant capabilities and attributes that can be reflected in each emergent phenotype: hunter-gatherer band, tool chain, business, state. 

    society
    good & bad
    The complexity of behavior is explored through Sapolsky developing scenarios of our best and worst behaviors across time spans, and scientific subjects including: anthropology, psychology, neuroscience, sociology.  The rich network of adaptive flows he outlines provides insights and highlight challenges for scientific research on behavior. 

    Complex adaptive system (CAS) theory builds on Sapolsky's details highlighting the strategies that evolution has captured to successfully enter niches we now occupy. 

    behaviors
    developed: caring for offspring, social cooperation, fighting over territory, capturing females, stealing resources; adaptations in evolutionary biology is a trait that increased the number of surviving offspring in an organism's ancestral lineage.  Holland argues: complex adaptive systems (CAS) adapt due to the influence of schematic strings on agents.  Evolution indicates fitness when an organism survives and reproduces.  For his genetic algorithm, Holland separated the adaptive process into credit assignment and rule discovery.  He assigned a strength to each of the rules (alternate hypothesis) used by his artificial agents, by credit assignment - each accepted message being paid for by the recipient, increasing the sender agent's rule's strength (implicit modeling) and reducing the recipient's.  When an agent achieved an explicit goal they obtained a final reward.  Rule discovery used the genetic algorithm to select strong rule schemas from a pair of agents to be included in the next generation, with crossing over and mutation applied, and the resulting schematic strategies used to replace weaker schemas.  The crossing over genetic operator is unlikely to break up a short schematic sequence that provides a building block retained because of its 'fitness';  In Deacon's conception of evolution, an adaptation is the realization of a set of constraints on candidate mechanisms, and so long as these constraints are maintained, other features are arbitrary. 
    of gene-culture co-evolution.  But our technologies can now alter and destroy the whole planet, making these attributes an existential threat to humanity.  To stress a point, Hawkins oversimplifies a complex, M. Mitchell Waldrop describes a vision of complexity via:
    • Rich interactions that allow a system to undergo spontaneous self-organization and, for CAS, evolution
    • Systems that are adaptive
    • More predictability than chaotic systems by bringing order and chaos into
    • Balance at the edge of chaos
    problem, explaining our brains help us occupy more niches and breed more children which has resulted in
    E. O. Wilson reviews the effect of man on the natural world to date and explains how the two systems can coexist most effectively. 
    8 billion people overloading the planet
    .  Our intelligence enables the achievement of goals in the face of obstacles.  The goals are sub-goals of genes' survival and reproduction and include:
    • Obtaining and eating food
    • Sex
    • Finding and maintaining shelter
    • Fighting for resources - in the preferred hunter-gatherer environment loss of resources was critical while possession was often transient. 
    • Understanding the proximate environment
    • Securing the cooperation of others 
    also
    The agents in complex adaptive systems (CAS) must model their environment to respond effectively to it.  Evolution's schematic operators and Samuel modeling together support the indirect recording of past successes and their strategic use by the current agent to learn how to succeed in the proximate environment. 
    models
    and predicts some of the consequences of uncontrolled population growth.  Yet we have continued to breed and stress the planet because the old brain is still in charge.  Hawkins notes that controlling population is socially and politically unacceptable, but he sees birth control pills as advanced technologies that can
    Terrence Deacon explores how constraints on dynamic flows can induce emergent phenomena which can do real work.  He shows how these phenomena are sustained.  The mechanism enables the development of Darwinian competition. 
    constrain
    the old brain: population growth could be solved by allowing women to control their own fertility.  He notes this strategy is clearly being resisted. 
  2. The neocortex is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 
    can be fooled into acting against its own best interests.  Hawkins expresses surprise that many people are comfortable accepting confusion and inconsistency in what they believe.  He concludes this is because:
    1. Many ideas can not be directly validated
    2. Contrary evidence is ignored
    3. The beliefs are incorrect meme-structured models; and he notes this framework matches the attributes of beliefs such as: vaccines are a core strategy of public health and have significantly extended global wellbeing over 200 years.  Smallpox & polio were virtually eradicated.  Recent successes include: HPV vaccine: Gardasil.  They induce active acquired immunity to a particular disease.  But the development and deployment of vaccines is complex:
      • The business model for vaccine development has been failing (Aug 2015): 
        • No Zika vaccine was available as the epidemic grew (Mar 2016).  No vaccine for: CMV;
        • Major foundations: Michael J. Fox, Gates, Wellcome; are working to improve the situation including sponsorship of the GAVI alliance.  A geographic cluster is forming in Seattle including PATH (Apr 2016). 
        • Commercial developers include: Affiris, Cell Genesis, Chiron, CSL, Sanofi, Valeant;
      • Vaccine deployment traditionally benefited from centrally managed vertical health programs.  But political issues are now constraining success with less than 95-99% coverage required for herd immunity (Aug 2015, Sep 2015, Nov 2015, Nov 2016, Jul 2018).  
        • Where clinics have been driven into local neighborhoods health improves (Apr 2016).  
        • Retail clinics (Mar 2016): CVS Minute Clinics focus on vaccination. 
        • NNT is a useful metric for vaccine benefit.  Influenza vaccine has an NNT of between 37 and 77, is cheap and causes little harm, so it is very beneficial. 
      • Key vaccines include: BCG, C. difficile (May 2015), Cholera (El Tor), Cervical Cancer (Gardasil HPV Jun 2018, Oct 2018), Dengvaxia (Mexico Dec 2015), Gvax, Influenza, Malaria vaccine, Provenge, Typbar-TCV (XDR typhoid Pakistan Apr 2018);
      • Regulation involves: FDA (CBER), with CMS monitoring (star ratings, PACE (Aug 2016), Report cards (Sep 2015)) & CDC promoting vaccines: as a sepsis measure, To control C. difficile (May 2015);  
        • Coding : CVX, MVX;
      • Research on vaccines includes: 
        • NIH: AIDS vaccines (AVRC), Focus on using genetic analysis to improve vaccine response.  
          • NCI:
            • Roswell Park clinical trial of immuno-oncology vaccine cimavax. 
        • Geisinger: effective process leverage in treatment. 
        • Stanford Edge immuno-oncology for cancer vaccines.  
        • P53-driven-cancer focused, gene therapy (Jun 2015). 
      cause autism is a major hereditary mental disorder that starts before age three when it features: a strong preference to be alone, a desire for things to stay the same, and areas of creative ability - they see the ordinary as beautiful and have special talents for: poetry, foreign languages, music, art, and calculations.  They generate less but more original ideas.  It occurs as a spectrum of symptoms, from mild to severe, across the population of sufferers (ASD).  Before age two the circumference of an autistic child's head is larger than typical and regions: amygdala, frontal lobe; develop prematurely, altering activity in other regions.  Autism highlights aspects of the brain's specialized regions and processes for interacting with other people.  Autistic's interests are restricted.  They struggle with social interactions & verbal and nonverbal communications.  Autistics do not attribute minds to other people: attributing mental states to others allows us to predict their behavior; a critical skill for social learning and interaction.  While their visual area MT detects motion, the superior temporal sulcus does not respond to biological motion in autistics, undermining the understanding of intention.  And they gaze at mouths rather than eyes when looking at faces.  The default mode network is disrupted.  Autistic adolescents have unusually large numbers of synapses, because of a failure of synaptic pruning.  Autistics almost never pretend.  They can't explain the difference between an instance of an object and a memory of it.  Mild autism still maintains some pressure to conform socially and often results in depression and anxiety.  Autism occurs in every country and social class.  It lasts a lifetime.  It has genetic and neurological causes.  Identical twins are 90% likely to both have autism if one of them does.  With 50% of genes active in the brain, mutations are likely to impact the development and operation of the brain.  The genes: SHANK3, CDH10; are involved but account for a very small percentage of the risk.  Facial gaze studies indicate a high genetic influence and an opportunity to identify more genes associated with autism (Jul 2017).  Copy number variations: an extra copy of a segment of 25 genes of chromosome 7 increases the risk of ASD, while deletion of the segment causes Williams syndrome; and de novo mutations which drive up the number of autism cases as paternal age has increased in the US.  ASD is associated with a reduced fusiform face area response.  Tests [in development] for autism include: SynapDx's blood test. 
      , climate change is not a threat, & There is an afterlife
Human intelligence, based on the old brain and neocortex while catalyzing, an infrastructure amplifier. 
the success of our species may also be the seed of our demise.  The old brain is in charge and the neocortex supports false beliefs.  Hawkins proceeds to explore ways out of this dilemma: simulation & merging, preserving knowledge, prophylactic actions

199 Merging brains and machines
Hawkins explains that scientists and technologists have started to seriously consider: uploading brains into computers, and merging our brains with computers; with some aiming to make it reality.  He reviews the two ideas framed by his model of the brain. 

Uploading a brain into a computer, seeks to simulate 'you' in a computer, which could be remote from earth, offering the possibility of avoiding the problem of being trapped on an uninhabitable Earth
The offer of merging a brain with a computer provides the potential to access Internet resources just by thinking.  This may help mitigate the feared intelligence explosion.  While the offers sound enticing, Hawkins is dubious about either possibility: simulation, merging

Hawkins notes that our bodies are programmed to die, while our brain is programmed to fear death.  He rationalizes the tension through the genes drive for a body to live long enough to reproduce copies of the genes.  For the body to live much longer, in a world of limited resources, may not improve the chances for the persistence of the replicator is Richard Dawkin's name for the genotype since it has the evolutionary goal of surviving long enough to reproduce its schematic plan effectively.  The action of genetic operators means that the results of successful reproduction may be different to the parental genotypes and phenotypes (Dawkin's vehicle). 
, now represented in the next generation.  But for Hawkins the
This page discusses the mechanisms and effects of emergence underpinning any complex adaptive system (CAS).  Physical forces and constraints follow the rules of complexity.  They generate phenomena and support the indirect emergence of epiphenomena.  Flows of epiphenomena interact in events which support the emergence of equilibrium and autonomous entities.  Autonomous entities enable evolution to operate broadening the adjacent possible.  Key research is reviewed. 
emergence
of intelligence enables the achievement of goals in the face of obstacles.  The goals are sub-goals of genes' survival and reproduction and include:
  • Obtaining and eating food
  • Sex
  • Finding and maintaining shelter
  • Fighting for resources - in the preferred hunter-gatherer environment loss of resources was critical while possession was often transient. 
  • Understanding the proximate environment
  • Securing the cooperation of others 
may not be in the best interest of the replicator, because we now know what is going on, and we can alter the process or the genes. 

202 Uploading your brain
Hawkins concludes there is a battle between
This page reviews the implications of selection, variation and heredity in a complex adaptive system (CAS).  The mechanism and its emergence are discussed. 
evolution
, paired with genes replicating is Richard Dawkin's name for the genotype since it has the evolutionary goal of surviving long enough to reproduce its schematic plan effectively.  The action of genetic operators means that the results of successful reproduction may be different to the parental genotypes and phenotypes (Dawkin's vehicle). 
, and our subjective intelligence enables the achievement of goals in the face of obstacles.  The goals are sub-goals of genes' survival and reproduction and include:
  • Obtaining and eating food
  • Sex
  • Finding and maintaining shelter
  • Fighting for resources - in the preferred hunter-gatherer environment loss of resources was critical while possession was often transient. 
  • Understanding the proximate environment
  • Securing the cooperation of others 
.  Hawkins sees a desire to escape and live forever, which is encouraging ideas about uploading your brain.  But Hawkins warns it would not be a satisfactory strategy.  We are
This page reviews the implications of reproduction initially generating a single initialized child cell.  For multi-cellular organisms this 'cell' must contain all the germ-line schematic structures including for organelles and multi-generational epi-genetic state.  Any microbiome is subsequently integrated during the innovative deployment of this creative event.  Organisms with skeletal infrastructure cannot complete the process of creation of an associated adult mind, until the proximate environment has been sampled during development.  The mechanism and resulting strategic options are discussed. 
organisms
so the computer would probably have to simulate the brain and body fully.  And if that complex, M. Mitchell Waldrop describes a vision of complexity via:
  • Rich interactions that allow a system to undergo spontaneous self-organization and, for CAS, evolution
  • Systems that are adaptive
  • More predictability than chaotic systems by bringing order and chaos into
  • Balance at the edge of chaos
task was achieved, the computer would experience being you.  You would not live forever, and while still alive you would have to coexist with a clone, slowly drifting in its own direction.  The result does not seem very rewarding for the human who still dies, indeed Hawkins highlights it is analogous to having children with your past. 

206 Merging your brain with a computer
Hawkins explains that electrodes are used to connect the brain with a computer aiming to allow information to 'flow' across the interface.  Currently there are brain computer interfaces that allow control of prosthetic limbs.  But the merging strategy is seen as a response to the intelligence explosion threat.  Hawkins doubts it will achieve a full uniting of brain and machine, and if it does it will still have a biological brain and body that will die. 

And from Hawkins's perspective neither scenario copes with the existential risks facing humanity

209 Estate planning for humanity
Hawkins shifts his focus to preserving humanity's knowledge, even if at that point in the future, humanity, like the dinosaurs, and eventually the Sun, no longer exist.  He assumes other intelligent beings, either
This page reviews the implications of selection, variation and heredity in a complex adaptive system (CAS).  The mechanism and its emergence are discussed. 
evolving
from life forms now on earth or extraterrestrials would benefit.  Hawkins values letting them know we existed.  And hopes to convey our history and knowledge.  He sees great value in understanding and sharing
Joseph Tainter introduces the problem of collapse and then develops a theory of complexity and reviews prior theories of collapse of societies.  He then builds a general explanation of collapse and explains declining marginal returns in significant aspects of complex societies, and evaluates the theory by examining its applicability to historical examples.  He then subsumes other explanatory themes into his marginal returns logic and applies it to our current situation. 

Following our summary of his arguments, RSS frames these from the perspective of complex adaptive system (CAS) theory: CAS entities provide an effective emergence and collapse point.  The history of events which results in each emergence point Tainter reviews introduces constraints on the aggregate entity.  These constraints can help define the emergence and collapse point and remove inconsistencies from the analytic framework.  Tainter's economic framework, conforming to the equilibrium proposed by Walras and Jevons, can benefit from alignment with complexity economics. 
collapse scenarios
.  Hawkins explores three scenarios:
  1. Message in a bottle - Pioneer and Voyager series probes included symbolic details for others to find.  Breakthrough Starshot aims to power tiny spacecraft with lasers, aiming to reach Alpha Centauri's planets.  But they are slow and have limited likelihood of being discovered. 
  2. Leave the lights on - Sending a powerful signal would increase the likelihood of being discovered.  While many people advise against this, Hawkins is keen, advised by his experience: to be a successful entrepreneur you have to tell others what you are doing.  And you are lucky if anyone cares about your idea.  Especially if they have to travel vast distances to capture it.  And probably other similarly intelligent life forms will not overlap in
    Carlo Rovelli resolves the paradox of time. 
    Rovelli initially explains that low level physics does not include time:
    • A present that is common throughout the universe does not exist
    • Events are only partially ordered.  The present is localized
    • The difference between past and future is not foundational.  It occurs because of state that through our blurring appears particular to us
    • Time passes at different speeds dependent on where we are and how fast we travel
    • Time's rhythms are due to the gravitational field
    • Our quantized physics shows neither space nor time, just processes transforming physical variables. 
    • Fundamentally there is no time.  The basic equations evolve together with events, not things 
    Then he explains how in a physical world without time its perception can emerge:
    • Our familiar time emerges
      • Our interaction with the world is partial, blurred, quantum indeterminate
      • The ignorance determines the existence of thermal time and entropy that quantifies our uncertainty
      • Directionality of time is real but perspectival.  The entropy of the world in relation to us increases with our thermal time.  The growth of entropy distinguishes past from future: resulting in traces and memories
      • Each human is a unified being because: we reflect the world, we formed an image of a unified entity by interacting with our kind, and because of the perspective of memory
      • The variable time: is one of the variables of the gravitational field.  With our scale we don't register quantum fluctuations, making space-time appear determined.  At our speed we don't perceive differences in time of different clocks, so we experience a single time: universal, uniform, ordered; which is helpful to our decisions

    time
    with us.  So Hawkins suggests building a powerful message that will transmit for millions of years - so self-sustaining.  His model is a star.  One mechanism would deploy Sun blockers which would be detectable over huge distances and would imply someone put them there.  The point is, once we figure out ways to do it we can look to see if others are using the methods. 
  3. Wiki earth - Hawkins sees knowledge as rare and intends to try and preserve it.  If humanity becomes extinct detailed evidence will degrade over a million years, unless processes are setup to conserve it.  Hawkins wants to preserve our knowledge in a more permanent form.  He proposes to add robustness by duplicating a Wikipedia style archive on satellites that orbit the Sun.  He imagines being able to update the archive but not delete it.  Even when the electronics stop functioning the archive would remain so if discovered it could be retrieved. 

223 Genes versus knowledge
Hawkins stresses that the balance of power between the old brain and the neocortex has begun to shift.  Recently, we have solved mysteries that perplexed our ancestors and the pace of knowledge expansion is accelerating.  And the physical universe appears to obey a single set of laws.  But our brains may support
The complexity of behavior is explored through Sapolsky developing scenarios of our best and worst behaviors across time spans, and scientific subjects including: anthropology, psychology, neuroscience, sociology.  The rich network of adaptive flows he outlines provides insights and highlight challenges for scientific research on behavior. 

Complex adaptive system (CAS) theory builds on Sapolsky's details highlighting the strategies that evolution has captured to successfully enter niches we now occupy. 

behaviors
that will generate a
Joseph Tainter introduces the problem of collapse and then develops a theory of complexity and reviews prior theories of collapse of societies.  He then builds a general explanation of collapse and explains declining marginal returns in significant aspects of complex societies, and evaluates the theory by examining its applicability to historical examples.  He then subsumes other explanatory themes into his marginal returns logic and applies it to our current situation. 

Following our summary of his arguments, RSS frames these from the perspective of complex adaptive system (CAS) theory: CAS entities provide an effective emergence and collapse point.  The history of events which results in each emergence point Tainter reviews introduces constraints on the aggregate entity.  These constraints can help define the emergence and collapse point and remove inconsistencies from the analytic framework.  Tainter's economic framework, conforming to the equilibrium proposed by Walras and Jevons, can benefit from alignment with complexity economics. 
societal collapse
, or total extinction.  Hawkins reviews three methods of preventing our demise:
  1. Become a multi-planet species - there are a variety of catastrophic events that Hawkins proposes to mitigate by becoming a two planet-species.  This is a justification for space flights to support a colony on Mars.  But Hawkins admits Mars is far from perfect.  He would mitigate the difficulties with intelligent autonomous robots, powered by the Sun, to manage the construction and maintenance of the operation.  For Hawkins the key is for the robots to leverage a neocortex is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 
    like feature.  He laments that even if the robots and Martian infrastructure was setup the human colony would eventually get into conflict with their neighbors back on Earth.  So he is keen to explore how to speed up space travel so more distant options could be tried to increase the robustness of the multi-planet strategy. 
  2. Modify our genes - Hawkins sees an option to give the neocortex is the main part of the cerebral cortex in mammals.  It is notable for its six layer structure and its layout as a matrix of cortical columns, a focus of Vernon Mountcastle and Jeff Hawkins.  It was originally thought to exist only in mammals but is also present in reptiles and birds buried behind other areas of the for-brain.  The for-brain develops based on a genetic plan consistent across all vertebrates.  The neocortex processes vision in the visual hierarchy V1, V2, V3 .. V5 ... V20; and language with areas including Wernicke's and Broca's with sensors in the inner ear.  Primate species with bigger social groups have larger cortices.  Human cortex size suggests traditional human cultures had an average size of 150 people. 
    , and its intelligence and desire to understand the world, an edge over the old brain and the process of
    This page reviews the implications of selection, variation and heredity in a complex adaptive system (CAS).  The mechanism and its emergence are discussed. 
    evolution
    .  Hawkins's proposed method is to alter our genes is clustered replicating interspersed silent palindromic repeats; a technique for exact targeting, cutting and editing of DNA based on leveraging bacterial enzymatic defenses against viruses generalized to any DNA sequence in a prokaryotic or eukaryotic cell.  It was identified during studies of a bacterial adaptive immune system.  In that system bacterial proteins grab parts of a virus that has infected them and record it within the palindromic structures that mark an array of inserted viral DNA used as a log persisted over generations.  If a new infection occurs the viral DNA is compared with the sequences and if a match exists the CAS proteins break up the viral DNA initiating its destruction.  This bacterial system was then updated and repurposed by the researchers to support targeted genetic engineering.  As explained by Dr. Doudna, the CRISPR proteins and the 20 nucleotide RNA template migrate into the nucleus where they rapidly target DNA which complements the RNA template and the Cas9 enzyme performs the edits.  Being a bacterial system CRISPR Cas9 does not target eukaryotic heterochromatic DNA well.  It is not fully understood how they find the target sequence so quickly.  It has been shown that Cas9 will bind to sites with a 5-8 base match but then it releases rapidly without cutting.  To cut, Cas9 has to reconfigure, which does not occur in the mismatch situations.   to "improve is, according to Dawkins, a suitcase word with at least five meanings in biology:
    1. Darwin and Wallace thought in terms of the capacity to survive and reproduce, but they were considering discrete aspects such as chewing grass - where hard enamel would improve the relative fitness. 
    2. Population geneticists: Ronald Fisher, Sewall Wright, J.B.S. Haldane; consider selection at a locus where for a genotype: green eyes vs blue eyes; one with higher fitness can be identified from genotypic frequencies and gene frequencies, with all other variations averaged out. 
    3. Whole organism 'integrated' fitness.  Dawkins notes there is only ever one instance of a specific organism.  Being unique, comparing the relative success of its offspring makes little sense.  Over a huge number of generations the individual is likely to have provided a contribution to everyone in the pool or no one. 
    4. Inclusive fitness, where according to Hamilton, fitness depends on an organism's actions or effects on its children or its relative's children, a model where natural selection favors organs and behaviors that cause the individual's genes to be passed on.  It is easy to mistakenly count an offspring in multiple relative's fitness assessments. 
    5. Personal fitness represents the effects a person's relatives have on the individual's fitness [3].  When interpreted correctly fitness [4] and fitness [5] are the same. 
    " our progeny to try and avoid going extinct, but he admits it may be impossible in principle
  3. Leave
    Desmond & Moore paint a picture of Charles Darwin's life, expanded from his own highlights:
    • His naughty childhood, 
    • Wasted schooldays,
    • Apprenticeship with Grant,
    • His extramural activities at Cambridge, walks with Henslow, life with FitzRoy on the Beagle,
    • His growing love for science,
    • London: geology, journal and Lyell. 
    • Moving from Gower Street to Down and writing Origin and other books. 
    • He reviewed his position on religion: the long dispute with Emma, his slow collapse of belief - damnation for unbelievers like his father and brother, inward conviction being evolved and unreliable, regretting he had ignored his father's advice; while describing Emma's side of the argument.  He felt happy with his decision to dedicate his life to science.  He closed by asserting after Self & Cross-fertilization his strength will be exhausted.  
    Following our summary of their main points, RSS frames the details from the perspective of complex adaptive system (CAS) theory.  Darwin placed evolution within a CAS framework, and built a network of supporters whose complementary skills helped drive the innovation. 
     
    Darwin
    's Orbit - aims to create machines that are intelligent and not dependent on us.  Hawkins highlights the opportunity: Preserve knowledge, Acquire new knowledge; which he sees as more directed than the
    This page reviews the implications of selection, variation and heredity in a complex adaptive system (CAS).  The mechanism and its emergence are discussed. 
    evolutionary
    mechanism with little direction and short term goals operating on molecules that replicate.  Hawkins concludes we can continue to live on Earth, working on keeping it a sustaining environment and limiting the impact of our worst behaviors.  And simultaneously dedicate resources to preservation of knowledge and supporting intelligence even when we no longer exist on Earth; with which he aims to catalyze, an infrastructure amplifier. 
    a discussion.   

241 Final thoughts
Hawkins hopes that understanding how the brain operates will support new capabilities in: medicine and mental health, true machine intelligence - which he sees as hugely beneficial, better ways to
Salman Khan argues that the evolved global education system is inefficient and organized around constraining and corralling students into accepting dubious ratings that lead to mundane roles.  He highlights a radical and already proven alternative which offers effective self-paced deep learning processes supported by technology and freed up attention of teams of teachers.  Building on his personal experience of helping overcome the unjustified failing grade of a relative, Khan:
  • Iteratively learns how to teach: Starting with Nadia, Leveraging short videos focused on content, Converging on mastery, With the help of neuroscience, and filling in dependent gaps; resulting in a different approach to the mainstream method. 
  • Assesses the broken US education system: Set in its ways, Designed for the 1800s, Inducing holes that are hidden by tests, Tests which ignore creativity.  The resulting teaching process is so inefficient it needs to be supplemented with homework.  Instead teachers were encouraging their pupils to use his tools at home so they could mentor them while they attended school, an inversion that significantly improves the economics. 
  • Enters the real world: Builds a scalable service, Working with a real classroom, Trying stealth learning, At Khan Academy full time,  In the curriculum at Los Altos, Supporting life-long learning. 
  • Develops The One World Schoolhouse: Back to the future with a one room school, a robust teaching team, and creativity enabled; so with some catalysis even the poorest can become educated and earn credentials for current jobs. 
  • Wishes he could also correct: Summer holidays, Transcript based assessments, College education;
  • Concludes it is now possible to provide the infrastructure for creativity to emerge and to support risk taking. 

Following our summary of his arguments RSS frames them from the perspective of complex adaptive system (CAS) theory.  Disruption is a powerful force for change but if its force is used to support the current teachers to adopt new processes can it overcome the extended phenotypic alignment and evolutionary amplifiers sustaining the current educational network? 

educate children
; all illustrations of intelligence enables the achievement of goals in the face of obstacles.  The goals are sub-goals of genes' survival and reproduction and include:
  • Obtaining and eating food
  • Sex
  • Finding and maintaining shelter
  • Fighting for resources - in the preferred hunter-gatherer environment loss of resources was critical while possession was often transient. 
  • Understanding the proximate environment
  • Securing the cooperation of others 
.  Hawkins sees understanding how the brain creates intelligence as the highest priority. 

Hawkins hopes readers will act on the ideas he has raised in the book: build a career in neuroscience or machine intelligence, help move these fields forward; enabling everyone to understand how their brain works and the implications.  Hawkins is convinced that everyone must understand the details: What makes us intelligent and self-aware? How did we become intelligent? What is the destiny of intelligence and knowledge?;  He suggests doing so will reduce conflicts and risk, is an assessment of the likelihood of an independent problem occurring.  It can be assigned an accurate probability since it is independent of other variables in the system.  As such it is different from uncertainty. 
of disaster.  And he hopes it will encourage us all to prioritize preserving intelligence and knowledge. 


Complex adaptive system (CAS)
This page introduces the complex adaptive system (CAS) theory frame.  The theory provides an organizing framework that is used by 'life.'  It can illuminate and clarify complex situations and be applied flexibly.  It can be used to evaluate and rank models that claim to describe our perceived reality.  It catalogs the laws and strategies which underpin the operation of systems that are based on the interaction of emergent agents.  It highlights the constraints that shape CAS and so predicts their form.  A proposal that does not conform is wrong. 

John Holland's framework for representing complexity is outlined.  Links to other key aspects of CAS theory discussed at the site are presented. 
theory
provides an organizing framework for
The agents in complex adaptive systems (CAS) must model their environment to respond effectively to it.  Evolution's schematic operators and Samuel modeling together support the indirect recording of past successes and their strategic use by the current agent to learn how to succeed in the proximate environment. 
modeling
This page discusses the mechanisms and effects of emergence underpinning any complex adaptive system (CAS).  Physical forces and constraints follow the rules of complexity.  They generate phenomena and support the indirect emergence of epiphenomena.  Flows of epiphenomena interact in events which support the emergence of equilibrium and autonomous entities.  Autonomous entities enable evolution to operate broadening the adjacent possible.  Key research is reviewed. 
emergent
capabilities such as Hawkins studies:


Hawkins's A Thousand Brains illuminates the brain science which identifies how cortical column based
Plans are interpreted and implemented by agents.  This page discusses the properties of agents in a complex adaptive system (CAS). 
It then presents examples of agents in different CAS.  The examples include a computer program where modeling and actions are performed by software agents.  These software agents are aggregates. 
The participation of agents in flows is introduced and some implications of this are outlined. 
agents
create a
The agents in complex adaptive systems (CAS) must model their environment to respond effectively to it.  Evolution's schematic operators and Samuel modeling together support the indirect recording of past successes and their strategic use by the current agent to learn how to succeed in the proximate environment. 
model
of the world and identify,
This page discusses the interdependence of perception and representation in a complex adaptive system (CAS).  Hofstadter and Mitchell's research with Copycat is reviewed.  The bridging of a node from a network of 'well known' percepts to a new representational instance is discussed as it occurs in biochemistry, within the brain, in consciousness and abstractly. 
represent
, and locate proximate objects is a collection of: happenings, occurrences and processes; including emergent entities, as required by relativity, explains Rovelli.  But natural selection has improved our fitness by representing this perception, in our minds, as an unchanging thing, as explained by Pinker.  Dehaene explains the object modeling and construction process within the unconscious and conscious brain.  Hawkins elaborates how the cortical columns of the neocortex identify and locate objects in our proximate environment and generate the stable representation.  Mathematicians view anything that can be defined and used in deductive reasoning and mathematical proofs as an object.  These mathematical objects can be values of variables, allowing them to be used in formulas.  
of interest. 






































































.
Market Centric Workshops
The Physics - Politics, Economics & Evolutionary Psychology
Politics, Economics & Evolutionary Psychology

Business Physics
Nature and nurture drive the business eco-system
Human nature
Emerging structure and dynamic forces of adaptation


integrating quality appropriate for each market
 
This page looks at schematic structures and their uses.  It discusses a number of examples:
  • Schematic ideas are recombined in creativity. 
  • Similarly designers take ideas and rules about materials and components and combine them. 
  • Schematic Recipes help to standardize operations. 
  • Modular components are combined into strategies for use in business plans and business models. 

As a working example it presents part of the contents and schematic details from the Adaptive Web Framework (AWF)'s operational plan. 

Finally it includes a section presenting our formal representation of schematic goals. 
Each goal has a series of associated complex adaptive system (CAS) strategy strings. 
These goals plus strings are detailed for various chess and business examples. 
Strategy
| Design |
This page uses an example to illustrate how:
  • A business can gain focus from targeting key customers,
  • Business planning activities performed by the whole organization can build awareness, empowerment and coherence. 
  • A program approach can ensure strategic alignment. 
Program Management
| Home

Profiles | Papers | Glossary | E-mail us