On Intelligence
Encyclopedia
On Intelligence: How a New Understanding of the Brain will Lead to the Creation of Truly Intelligent Machines is a book by Palm Pilot
-inventor Jeff Hawkins
with New York Times science writer Sandra Blakeslee. The book explains Hawkins' memory-prediction framework
theory of the brain
and describes some of its consequences. (Times Books: 2004, ISBN 0-8050-7456-2)
to illustrate a conflict of ideas. Hawkins believed (and ostensibly continues to believe) creating true artificial intelligence will only be possible with intellectual progress in the discipline of neuroscience. Hawkins writes that the scientific establishment (as symbolized by MIT) has historically rejected the relevance of neuroscience to artificial intelligence. Indeed, some artificial intelligence researchers have "[taken] pride in ignoring neurobiology" (p. 12).
Hawkins is an electrical engineer by training, and a neuroscientist by inclination. He used electrical engineering concepts as well as the studies of neuroscience to formulate his framework. In particular, Hawkins treats the propagation of nerve impulses in our nervous system
as an encoding problem
, specifically, a future predicting state machine, similar in principle to feed-forward error-correcting state machines.
.
The state machine actually controls the behavior of the organism. Since it is a feed forward state machine, the organism responds to future events predicted from past data.
The hierarchy is capable of memorizing frequently observed sequences (Cognitive modules) of patterns and developing invariant representations. Higher levels of the cortical hierarchy predict the future on a longer time scale, or over a wider range of sensory input. Lower levels interpret or control limited domains of experience, or sensory or effector systems. Connections from the higher level states predispose some selected transitions in the lower-level state machines.
Hebbian learning is part of the framework, in which the event of learning physically alters neurons and connections, as learning takes place.
Vernon Mountcastle
's formulation of a cortical column
is a basic element in the framework. Hawkins places particular emphasis on the role of the interconnections from peer columns, and the activation of columns as a whole. He strongly implies that a column is the cortex's physical representation of a state in a state machine.
As an engineer, any specific failure to find a natural occurrence of some process in his framework does not signal a fault in the memory-prediction framework per se, but merely signals that the natural process has performed Hawkins' functional decomposition in a different, unexpected way, as Hawkins' motivation is to create intelligent machine
s. For example, for the purposes of his framework, the nerve impulses can be taken to form a temporal sequence (but phase encoding could be a possible implementation of such a sequence; these details are immaterial for the framework).
s use the visual system
as a prototype for some example predictions, such as Predictions 2, 8, 10, and 11. Other predictions cite the auditory system
( Predictions 1, 3, 4, and 7).
, Hawkins (2004) predicts "we should find anticipatory cells", cells that fire in anticipation of a sensory event
.
, Hawkins predicts, for example, "we should find anticipatory cells in or near V1, at a precise location in the visual field (the scene)". It has been experimentally determined, for example, after mapping the angular position of some objects in the visual field, there will be a one-to-one correspondence of cells in the scene to the angular positions of those objects. Hawkins predicts that when the features of a visual scene are known in a memory, anticipatory cells should fire before the actual objects are seen in the scene.
, the repository of new memories.
of the eye
over a learned scene, for example).
s should be capable of detecting coincident events on thin dendrite
s, even for a neuron
with thousands of synapse
s. Hawkins posits a temporal window (presuming time-encoded firing) which is necessary for his theory
to remain viable.
Palm (PDA)
Palm handhelds were Personal Digital Assistants which ran the Palm OS. Palm devices have evolved from handhelds to smartphones which run Palm OS, WebOS, and Windows Mobile...
-inventor Jeff Hawkins
Jeff Hawkins
Jeffrey Hawkins is the founder of Palm Computing and Handspring...
with New York Times science writer Sandra Blakeslee. The book explains Hawkins' memory-prediction framework
Memory-prediction framework
The memory-prediction framework is a theory of brain function that was created by Jeff Hawkins and described in his 2004 book On Intelligence...
theory of the brain
Brain
The brain is the center of the nervous system in all vertebrate and most invertebrate animals—only a few primitive invertebrates such as sponges, jellyfish, sea squirts and starfishes do not have one. It is located in the head, usually close to primary sensory apparatus such as vision, hearing,...
and describes some of its consequences. (Times Books: 2004, ISBN 0-8050-7456-2)
Outline
Hawkins outlines the book as follows:A personal history
The first chapter is a brief history of Hawkins' interest in neuroscience juxtaposed against a history of artificial intelligence research. Hawkins uses a story of his failed application to the Massachusetts Institute of TechnologyMassachusetts Institute of Technology
The Massachusetts Institute of Technology is a private research university located in Cambridge, Massachusetts. MIT has five schools and one college, containing a total of 32 academic departments, with a strong emphasis on scientific and technological education and research.Founded in 1861 in...
to illustrate a conflict of ideas. Hawkins believed (and ostensibly continues to believe) creating true artificial intelligence will only be possible with intellectual progress in the discipline of neuroscience. Hawkins writes that the scientific establishment (as symbolized by MIT) has historically rejected the relevance of neuroscience to artificial intelligence. Indeed, some artificial intelligence researchers have "[taken] pride in ignoring neurobiology" (p. 12).
Hawkins is an electrical engineer by training, and a neuroscientist by inclination. He used electrical engineering concepts as well as the studies of neuroscience to formulate his framework. In particular, Hawkins treats the propagation of nerve impulses in our nervous system
Nervous system
The nervous system is an organ system containing a network of specialized cells called neurons that coordinate the actions of an animal and transmit signals between different parts of its body. In most animals the nervous system consists of two parts, central and peripheral. The central nervous...
as an encoding problem
Code
A code is a rule for converting a piece of information into another form or representation , not necessarily of the same type....
, specifically, a future predicting state machine, similar in principle to feed-forward error-correcting state machines.
The theory
Hawkins' basic idea is that the brain is a mechanism to predict the future, specifically, hierarchical regions of the brain predict their future input sequences. Perhaps not always far in the future, but far enough to be of real use to an organism. As such, the brain is a feed forward hierarchical state machine with special properties that enable it to learnLearning theory
Learning theory may refer to:* Learning theory , the process of how humans learn** Behaviorism** Cognitivism** Constructivism** Connectivism* Computational learning theory, a mathematical theory to analyze machine learning algorithms...
.
The state machine actually controls the behavior of the organism. Since it is a feed forward state machine, the organism responds to future events predicted from past data.
The hierarchy is capable of memorizing frequently observed sequences (Cognitive modules) of patterns and developing invariant representations. Higher levels of the cortical hierarchy predict the future on a longer time scale, or over a wider range of sensory input. Lower levels interpret or control limited domains of experience, or sensory or effector systems. Connections from the higher level states predispose some selected transitions in the lower-level state machines.
Hebbian learning is part of the framework, in which the event of learning physically alters neurons and connections, as learning takes place.
Vernon Mountcastle
Vernon Mountcastle
Vernon Benjamin Mountcastle is Professor Emeritus of Neuroscience at Johns Hopkins University.He discovered and characterized the columnar organization of the cerebral cortex in the 1950s...
's formulation of a cortical column
Cortical column
A cortical column, also called hypercolumn or sometimes cortical module, is a group of neurons in the brain cortex which can be successively penetrated by a probe inserted perpendicularly to the cortical surface, and which have nearly identical receptive fields...
is a basic element in the framework. Hawkins places particular emphasis on the role of the interconnections from peer columns, and the activation of columns as a whole. He strongly implies that a column is the cortex's physical representation of a state in a state machine.
As an engineer, any specific failure to find a natural occurrence of some process in his framework does not signal a fault in the memory-prediction framework per se, but merely signals that the natural process has performed Hawkins' functional decomposition in a different, unexpected way, as Hawkins' motivation is to create intelligent machine
Machine
A machine manages power to accomplish a task, examples include, a mechanical system, a computing system, an electronic system, and a molecular machine. In common usage, the meaning is that of a device having parts that perform or assist in performing any type of work...
s. For example, for the purposes of his framework, the nerve impulses can be taken to form a temporal sequence (but phase encoding could be a possible implementation of such a sequence; these details are immaterial for the framework).
Predictions of the theory of the memory-prediction framework
His predictionPrediction
A prediction or forecast is a statement about the way things will happen in the future, often but not always based on experience or knowledge...
s use the visual system
Visual system
The visual system is the part of the central nervous system which enables organisms to process visual detail, as well as enabling several non-image forming photoresponse functions. It interprets information from visible light to build a representation of the surrounding world...
as a prototype for some example predictions, such as Predictions 2, 8, 10, and 11. Other predictions cite the auditory system
Auditory system
The auditory system is the sensory system for the sense of hearing.- Outer ear :The folds of cartilage surrounding the ear canal are called the pinna...
( Predictions 1, 3, 4, and 7).
- An Appendix of 11 Testable Predictions:
Enhanced neural activity in anticipation of a sensory event
1. In all areas of cortexCerebral cortex
The cerebral cortex is a sheet of neural tissue that is outermost to the cerebrum of the mammalian brain. It plays a key role in memory, attention, perceptual awareness, thought, language, and consciousness. It is constituted of up to six horizontal layers, each of which has a different...
, Hawkins (2004) predicts "we should find anticipatory cells", cells that fire in anticipation of a sensory event
Phenomenon
A phenomenon , plural phenomena, is any observable occurrence. Phenomena are often, but not always, understood as 'appearances' or 'experiences'...
.
- Note: As of 2005 mirror neuronMirror neuronA mirror neuron is a neuron that fires both when an animal acts and when the animal observes the same action performed by another. Thus, the neuron "mirrors" the behaviour of the other, as though the observer were itself acting. Such neurons have been directly observed in primate and other...
s have been observed to fire before an anticipated event.
Spatially specific prediction
2. In primary sensory cortexCerebral cortex
The cerebral cortex is a sheet of neural tissue that is outermost to the cerebrum of the mammalian brain. It plays a key role in memory, attention, perceptual awareness, thought, language, and consciousness. It is constituted of up to six horizontal layers, each of which has a different...
, Hawkins predicts, for example, "we should find anticipatory cells in or near V1, at a precise location in the visual field (the scene)". It has been experimentally determined, for example, after mapping the angular position of some objects in the visual field, there will be a one-to-one correspondence of cells in the scene to the angular positions of those objects. Hawkins predicts that when the features of a visual scene are known in a memory, anticipatory cells should fire before the actual objects are seen in the scene.
Prediction should stop propagating in the cortical column at layers 2 and 3
3. In layers 2 and 3, predictive activity (neural firing) should stop propagating at specific cells, corresponding to a specific prediction. Hawkins does not rule out anticipatory cells in layers 4 and 5."Name cells" at layers 2 and 3 should preferentially connect to layer 6 cells of cortex
4. Learned sequences of firings comprise a representation of temporally constant invariants. Hawkins calls the cells which fire in this sequence "name cells". Hawkins suggests that these name cells are in layer 2, physically adjacent to layer 1. Hawkins does not rule out the existence of layer 3 cells with dendrites in layer 1, which might perform as name cells."Name cells" should remain ON during a learned sequence
5. By definition, a temporally constant invariant will be active during a learned sequence. Hawkins posits that these cells will remain active for the duration of the learned sequence, even if the remainder of the cortical column is shifting state. Since we do not know the encoding of the sequence, we do not yet know the definition of ON or active; Hawkins suggests that the ON pattern may be as simple as a simultaneous AND (i.e., the name cells simultaneously "light up") across an array of name cells.- See Neural ensemble#Encoding for grandmother neurons which perform this type of function.
"Exception cells" should remain OFF during a learned sequence
6. Hawkins' novel prediction is that certain cells are inhibited during a learned sequence. A class of cells in layers 2 and 3 should NOT fire during a learned sequence, the axons of these "exception cells" should fire only if a local prediction is failing. This prevents flooding the brain with the usual sensations, leaving only exceptions for post-processing."Exception cells" should propagate unanticipated events
7. If an unusual event occurs (the learned sequence fails), the "exception cells" should fire, propagating up the cortical hierarchy to the hippocampusHippocampus
The hippocampus is a major component of the brains of humans and other vertebrates. It belongs to the limbic system and plays important roles in the consolidation of information from short-term memory to long-term memory and spatial navigation. Humans and other mammals have two hippocampi, one in...
, the repository of new memories.
"Aha! cells" should trigger predictive activity
8. Hawkins predicts a cascade of predictions, when recognition occurs, propagating down the cortical column (with each saccadeSaccade
A saccade is a fast movement of an eye, head or other part of an animal's body or device. It can also be a fast shift in frequency of an emitted signal or other quick change. Saccades are quick, simultaneous movements of both eyes in the same direction...
of the eye
Human eye
The human eye is an organ which reacts to light for several purposes. As a conscious sense organ, the eye allows vision. Rod and cone cells in the retina allow conscious light perception and vision including color differentiation and the perception of depth...
over a learned scene, for example).
Pyramidal cells should detect coincidences of synaptic activity on thin dendrites
9. Pyramidal cellPyramidal cell
Pyramidal neurons are a type of neuron found in areas of the brain including cerebral cortex, the hippocampus, and in the amygdala. Pyramidal neurons are the primary excitation units of the mammalian prefrontal cortex and the corticospinal tract. Pyramidal neurons were first discovered and...
s should be capable of detecting coincident events on thin dendrite
Dendrite
Dendrites are the branched projections of a neuron that act to conduct the electrochemical stimulation received from other neural cells to the cell body, or soma, of the neuron from which the dendrites project...
s, even for a neuron
Neuron
A neuron is an electrically excitable cell that processes and transmits information by electrical and chemical signaling. Chemical signaling occurs via synapses, specialized connections with other cells. Neurons connect to each other to form networks. Neurons are the core components of the nervous...
with thousands of synapse
Synapse
In the nervous system, a synapse is a structure that permits a neuron to pass an electrical or chemical signal to another cell...
s. Hawkins posits a temporal window (presuming time-encoded firing) which is necessary for his theory
Theory
The English word theory was derived from a technical term in Ancient Greek philosophy. The word theoria, , meant "a looking at, viewing, beholding", and referring to contemplation or speculation, as opposed to action...
to remain viable.
Learned representations move down the cortical hierarchy, with training
10. Hawkins posits, for example, that if the inferotemporal (IT) layer has learned a sequence, that eventually cells in V4 will also learn the sequence."Name cells" exist in all regions of cortex
11. Hawkins predicts that "name cells" will be found in all regions of cortex.See also
- Hierarchical temporal memoryHierarchical Temporal MemoryHierarchical temporal memory is a machine learning model developed by Jeff Hawkins and Dileep George of Numenta, Inc. that models some of the structural and algorithmic properties of the neocortex. HTM is a biomimetic model based on the memory-prediction theory of brain function described by Jeff...
, a technology by Hawkins's startup NumentaNumentaNumenta is a company founded March 24, 2005, by Palm founder Jeff Hawkins with his longtime business partner Donna Dubinsky and Stanford graduate student Dileep George. It is headquartered in Redwood City, California.-Origin:...
Inc. to replicate the properties of the neocortex. - Memory-prediction frameworkMemory-prediction frameworkThe memory-prediction framework is a theory of brain function that was created by Jeff Hawkins and described in his 2004 book On Intelligence...
External links
- OnIntelligence.com - official website
- A Hierarchical Bayesian Model of Invariant Pattern Recognition in the Visual Cortex, a conference paper by George & Hawkins
- Saulius Garalevicius' research page - Research papers and programs presenting experimental results with Bayesian models of the Memory-Prediction Framework
- Project Neocortex - An open source project for modeling Memory-Prediction Framework
Reviews
- Machine Intelligence Meets Neuroscience (By Bob Colwell, published in IEEE's Computer, January 2005)
- Above link is broken, although the Internet ArchiveInternet ArchiveThe Internet Archive is a non-profit digital library with the stated mission of "universal access to all knowledge". It offers permanent storage and access to collections of digitized materials, including websites, music, moving images, and nearly 3 million public domain books. The Internet Archive...
has a copy. - Machine Intelligence Meets Neuroscience (full text for IEEE Explore's online subscribers)
- Machine Intelligence Meets Neuroscience (citation only)
- Above link is broken, although the Internet Archive
- A review by Franz Dill
- On Intelligence, People and Computers (Arnold Kling, Tech Central Station, 22 November 2004)
- On Biological and Digital Intelligence A review by Ben GoertzelBen GoertzelBen Goertzel , is an American author and researcher in the field of artificial intelligence. He currently leads Novamente LLC, a privately held software company that attempts to develop a form of strong AI, which he calls "Artificial General Intelligence"...
(7 Oct 2004)