About
290
Publications
26,096
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
3,435
Citations
Introduction
Additional affiliations
Position
- Emeritus Professor and Senior Research Investigator
Publications
Publications (290)
Here is examined the work done in many laboratories on the proposition that the mechanisms underlying consciousness in living organisms can be studied using computational theories. This follows an agreement at a 2001 multi-disciplinary meeting of philosophers, neuroscientists and computer scientists that such a research programme was feasible and w...
As robots are generally thought to perform human-like tasks, they depend on the successes of information technology in the area of artificial intelligence to succeed in such pursuits. But robots, through their anthropomorphic character and their weighty presence in science fiction, attract the attention of the press and the media in a way that, at...
The first edition of the book was written to introduce curious readers to the way that the consciousness we all enjoy might depend on the networks of neurons that make up the brain. In this revised edition, it is recognized that these arguments still stand, but that they have been taken much further by an increasing number of researchers.Each chapt...
Five reviewers have kindly read our \Aristotle's Laptop: Discovering our Informational Minds."
Here, we thank them for spending their time, meet some of their comments and indicate
developments which their comments encourage.
A weightless neural state machine acting as an exploratory automaton changes its position in a simulated toy world by its own actions. A popular question is asked: how might the automaton ‘become conscious of’ the effect of its own actions? Here we develop previously defined iconic learning in such weightless machines so that this knowledge can be...
The stated aim of adherents to the paradigm called biologically inspired cognitive architectures (BICA) is to build machines that address "the challenge of creating a real-life computational equivalent of the human mind".(From the mission statement of the new BICA journal.) In contrast, practitioners of machine consciousness (MC) are driven by the...
Information integration is a measure, developed by Tononi and co-researchers, of the capacity for dynamic neural networks to be in informational states which are unique and indivisible. This is supposed to correspond to the intuitive "feel" of a mental state: highly discriminative and yet fundamentally integrated. Recent versions of the theory incl...
Machine consciousness has emerged from the confusion of an oxymoron
into an evolving set of principles which, by leaning on information integration
theories, define and distinguish what is meant by ‘a conscious machine’.
This paper reviews this process of emergence by indicating how it is possible to
break away from the Chalmers ‘hardness’ of a com...
Aristotle’s convincing philosophy is likely to have shaped
(even indirectly) many of our current beliefs, prejudices
and attitudes to life. This includes the way in which our mind
(that is, our capacity to have private thoughts) appears to
elude a scientific description. This book is about a scientific
ingredient that was not available to Aristotle...
Much discussion on the singularity is based on the assumption that the design ability of a human can be transferred into an AI system, then rendered autonomous and self-improving. I argue here that this cannot be foreseen from the current state of the art of automatic or evolutionary design. Assuming that this will happen 'some day' is a doubtful s...
We discuss the role of function application in neural models of visual awareness with reference to the handling of language understanding. The problem of modeling the application of functional relations mapping a set of given input variables to a defined output outcome is seen to arise in the context of enabling a visual awareness model to handle v...
A number of people have suggested that there is a link between information integration and consciousness, and a number of algorithms for calculating information integration have been put forward. The most recent of these is Balduzzi and Tononi's state-based Φ algorithm, which has factorial dependencies that severely limit the number of neurons that...
In this paper, we present a depictive neural model for the representation of motion verb semantics in neural models of visual awareness. The problem of modelling motion verb representation is shown to be one of function application, mapping a set of given input variables defining the moving object and the path of motion to a defined output outcome...
In recent years a number of people have suggested that there is a close link between conscious experience and the differentiation and integration of information in certain areas of the brain. The balance between differentiation and integration is often called information integration, and a number of algorithms for measuring it have been put forward...
We discuss a mechanism for the representation of location and path references by purely neural means in a digital model of
visual awareness. The underlying base awareness model features a “virtual eye” that can look out into a virtual world populated
with objects to be recognised. The presented models take in to account location information sensed...
Brain Inspired Cognitive Systems 2008 (June 24-27, 2008; São Luís, Brazil) brought together leading scientists and engineers who use analytic, syntactic and computational methods both to understand the prodigious processing properties of biological systems and, specifically, of the brain, and to exploit such knowledge to advance computational metho...
This paper aligns recent developments on information integration and consciousness with work on axiomatic systems. Axiomatic models are usually implemented as weightless neural structures. It is argued that weightless approaches lead to a logic-based interpretation of the information generated by the states of a network. Learning of phenomenal stat...
This paper critically tracks the impact of the development of the machine consciousness paradigm from the incredulity of the 1990s to the structuring of the turn of this century, and the consolidation of the present time which forms the basis for guessing what might happen in the future. The underlying question is how this development may have chan...
This paper reviews computational work that is currently developing under the heading of ‘Machine Consciousness’ and sets out
to provide a guide for those who wish to contribute to this field. First, questions of philosophical concern as to the appropriateness
of this activity are raised and discussed. Then some classical designs and computational a...
It is shown that the application of a form of spike time dependent plasticity (STDP) within a highly recurrent spiking neural net based upon the LSM leads to an approximate convergence of the synaptic weights. Convergence is a desirable property as it ...
This paper describes how early designs of dynamic weightless neural systems were developed to enable some of the states of a state structure to have a phenomenal character. Such states reflect the features of a sensory reality and allow the storage of aspects of sensory experience and access to it. The ‘machine consciousness ’ paradigm is summarise...
Mimicking biological neurons by focusing on the excitatory/inhibitory decoding performed by the dendritic trees is a different and attractive alternative to the integrate-and-fire McCullogh-Pitts neuron stylisation. In such alternative analogy, neurons can be seen as a set of RAM nodes addressed by Boolean inputs and producing Boolean outputs. The...
This paper argues that supervised cognitive growth in artifacts will be very difficult to achieve without detailed knowledge about systems" internal states. Physical information is too low level to provide a useful understanding of a system"s behavior, and it is more pragmatically useful to take a mental stance towards an artificial system and inte...
In discussions about the physical support of conscious experience, a recent trend has been introduced (by Tononi and various colleagues) that measures the capacity of a network to discriminate among different states and integrate the information generated by this discrimination. This capacity to generate and integrate information can be used to und...
‘Machine Consciousness’, which some years ago might have been suppressed as an inappropriate pursuit, has come out of the closet and is now a legitimate area of research concern. This paper briefly surveys the last few years of worldwide research in this area which divides into rule‐based and neural approaches and then reviews the work of the autho...
In this chapter we present a computational architecture intended to add clarity to the concept of consciousness. We briefly review some of the motivations of work done in this area in various institutes around the world and looks closely at our own work which specifically includes phenomenology, the sense of a self in a perceptual world. This break...
While the recent special issue of JCS on machine consciousness (Volume 14, Issue 7) was in preparation, a collection of papers on the same topic, entitled Artificial Consciousness and edited by Antonio Chella and Riccardo Manzotti, was published. The editors of the JCS special issue, Ron Chrisley, Robert Clowes and Steve Torrance, thought it would...
A model of conscious mechanisms called Axiomatic Consciousness Theory (ACT) is used to develop a neuroarchitectural model of visual phenomenology. The result is an extension of concepts in neural systems towards phenomenal intentionality: the lack of which is a common starting point for critiques of AI. Here the argument is developed at four intera...
This paper looks closely at previously enunciated axioms that specifically include phenomenology as the sense of a self in a perceptual world. This, we suggest, is an appropriate way of doing science on a first-person phenomenon. The axioms break consciousness down into five key components: presence, imagination, attention, volition and emotions. T...
Is synthetic phenomenology a valid concept? In approaching consciousness from a computational point of view, the question of phenomenology is not often explicitly addressed. In this paper we re- view the use of phenomenology as a philosophical and a cognitive construct in order to have a meaningful transfer of the concept into the computational dom...
This paper describes the efforts of those who work with informational machines and with informational analyses to provide a basis for understanding consciousness and for speculating on what it would take to make a conscious machine. Some of the origins of these considerations are covered and the contributions of several researchers are reviewed. A...
A development of dynamic neural networks has been to view them as learning state machines (Aleksander and Morton 1993, 1995 and Aleksander 1996). This paper formalises ideas previously presented as empirical possibilities in order to show how particular modes of learning influence the capacity of such systems to model their environments which too h...
In this paper we argue that phenomenology needs to be supported by explicit mechanisms if one is to have computational models of consciousness. Computational work in this area is reviewed and a set of axioms that help to decompose being conscious into manageable concepts is evoked. This leads to a kernel architecture and a digital implementation wh...
In recent years, issues of inattention blindness and change blindness have thrown doubt on theories of vision that assume
that the visual signal is inwardly represented for further recognition and processing. The aim of this paper is to review
so called enacted theories of vision and argue that they are too severe in terms of removing inner represe...
The work from several laboratories on the modeling of consciousness is reviewed. This ranges, on one hand, from purely functional models where behavior is important and leads to an attribution of consciousness to, on the other hand, material work closely derived from the information about the anatomy of the brain. At the functional end of the spect...
Never one to dodge the big questions, Igor Aleksander tackles them head on in this latest book: Is there a real world? or is our awareness of one an illusion? Are animals conscious? Can machines be conscious? What is it to be unconscious? Does the science of consciousness impinge on religious thought?
The views of professor Emiritus in neural system engineering, Igor Aleksander, on the development of robots, and mobile phones, which understand their owners, and machine consciousness programmes to be used in commercial applications and medical benefits, are discussed. Researchers are pursuing various design methods, and philosophies attempting to...
The way in which artificial intelligence has developed over the last 50 years has had a major role in shaping cognitive science as it is today. This has generated computational models of behaviour. The connectionist revival of the 1980s added a tinge of neurodynamics to this. Here I suggest that some post-connectionist work in artificial intelligen...
Is artificial intelligence (AI) just something that is done in laboratories disconnected from the development of the pragmatic computing, which constitutes current information technology or does it contribute to progress in computing and information technology? It has even been suggested that advances in AI are merely a re-branding exercise for pro...
Visual deficits have been discovered in sufferers of Parkinson’ disease [1] and the cause of this is not clearly understood. This paper reports on a digital neuromodel that investigates a hypothesis
that the deficit may be due to a projection from the Basal Ganglia to the Superior Colliculus where a shortage of Dopamine
introduces noise in the ocu...
A Generalising Random Access Memory (G-RAM) neuron is distinguished from conventional neuron models by the fact that its tolerance to departures in similarity from its training pattern is variable. Details of this are given in this paper as it affects the behaviour a class of digital probabilistic neural networks which have been achieving attention...
A new approach that identifies necessary design requirements for neural machinery that one could claim is conscious is presented.
Usually such designs stem from a third-person inference of suitable mechanisms which are thought to underpin conscious behaviour.
The novelty here comes from using the first person (introspection) to identify a set of ne...
This paper relates to a formal statement of the mechanisms that are thought minimally necessary to underpin consciousness. This is expressed in the form of axioms. We deem this to be useful if there is ever to be clarity in answering questions about whether this or the other organism is or is not conscious. As usual, axioms are ways of making forma...
Undoubtedly John Searle’s central contribution to the philosophy of both mind and language is his theory of intentionality. This suggests that cognition in all its forms is directed at objects. We attend to objects, we speak of objects, we understand object-laden stories, and we relate to objects both in perception and recall. When programmed compu...
Nature is the international weekly journal of science: a magazine style journal that publishes full-length research papers in all disciplines of science, as well as News and Views, reviews, news, features, commentaries, web focuses and more, covering all branches of science and how science impacts upon all aspects of society and life.
Interweaving anecdotes from his own life and research with imagined dialogues between historical figures -- including Descartes, Locke, Hume, Kant, Wittgenstein, Francis Crick, and Steven Pinker -- Aleksander leads readers toward an understanding of consciousness. He shows not only how the latest work with artificial neural systems suggests that an...
The object of seeing is for the brain to create inner states that accurately model the world and recall it for purposeful
use. In this descriptive paper we present virtual neuro-architectures called ‘depictive’ which have been developed to create
hypotheses for the mechanisms necessary for such depiction and explain some elements of verbally induce...
This work supports the hypothesis that behavior is the result of rich but mediated representations of the real world and the organism within it. ‘Rich’ means containing all the detail necessary for appropriate action taking and ‘mediated’ means that representations of ’self are at work which include decisions on whether and how to act. The central...
An existing hypothesis states that visual consciousness is made up of 'microconsciousnesses' occurring asynchronously in several sites of the visual system of the brain with no need for direct means of binding. We extend this hypothesis to define what qualifies a neural activity for generating an element of consciousness to distinguish it from one...
How can a system with visual input become capable of visualising what is meant by new combinations of known words? For example,
it is possible for most of us to visualise a blue banana with red spots even though such an object would never have formed
part of our experience. In this paper we discuss a neural system which is capable of simple forms o...
How can a system with visual input become capable of visualising what is meant by new combinations of known words? For example, it is possible for most of us to visualise a blue banana with red spots even though such an object would never have formed part of our experience. The authors discuss a neural system which is capable of simple forms of thi...
This chapter reviews a progression of weightless systems from the WISARD single-layer pattern recogniser to recent work on MAGNUS, a state machine designed to store sensory experience in its state structure. The stress is on algorithmic effects, that is, the effect of mapping these systems into conventional processors as virtual neural machines. Th...
This paper synthesises three diverse approaches to the study of consciousness in a description of an existing program of work in Artificial Neuroconsciousness. The three approaches are drawn from automata theory ([Aleksander, 1995][Aleksander, 1996]), psychology ([Karmiloff-Smith, 1992]; [Clark Karmiloff-Smith, 1993]) and philosophy ([Searle, 1992]...
A development of dynamic neural networks has been to view them as learning state machines (Aleksander and Morton 1993,1995 and Aleksander 1996). This paper formalises ideas previously presented as empirical possibilities in order to show how particular modes of learning influence the capacity of such systems to model their environments which too ha...
The VC dimension of an n-tuple classifier predicts poorer
generalisation performance than is found in practice. The effective VC
dimension of the classifier for a Gaussian input distribution is
calculated empirically and shown to be significantly lower than the true
VC dimension. Thus better generalisation can be predicted without the
risk of over-...
Arguing that the neurons in our heads are the source of consciousness, this book attempts to explain how this happens. Although it talks of neural networks, it tries to explain what they are and what they do, in such a way that anyone may understand. This book is also a story. A story of a land where people think that they are automata without much...
Impossible Minds: My Neurons, My Consciousness has been written to satisfy the curiosity each and every one of us has about our own consciousness. It takes the view that the neurons in our heads are the source of consciousness and attempts to explain how this happens. Although it talks of neural networks, it explains what they are and what they do...
The topic of artificial consciousness asks the following question. Were a manufactured object appear to be conscious what mechanisms could be responsible for such a state of affairs? First the paper defines the foundation of the theory as a postulate which requires the manufactured object to be neural, specifying that the firing of some neurons is...
A partially connected general neural unit, which is capable of
making state transitions with controlled probabilities, is described.
The conditions for which the cost of the partially connected system is
less than the cost of a fully connected network are found. The network
performance and optimisation are then confirmed experimentally
A Neural State Machine, constructed from weightless artificial neurons, is used to ground visual and linguistic representations, from real-world sources, using minimal preprocessing. Simulations are used to demonstrate that the system can work in real time, in a real environment, to provide a means of grounding both concrete nouns, and some simple...
The term “intentionality” arises in connection with
natural language understanding in a computer. The problem is not one of
speech recognition. It remains a problem even if the words of the
language were perfectly encoded by a speech recognizer, or even typed on
a keyboard. It is believed that the ability to visualize events when
hearing sentences...
The concept of a theory of artificial neural consciousness based on neural machines was introduced at ICANN94 (Aleksander, 1994). Here the theory is developed by defining that which would have to be synthesized were consciousness to be found in an engineered artefact. This is given the name artificial consciousness to indicate that the theory is ob...
A method to estimate how the Boolean space is distributed in relation to two patterns separated by a distance h is presented. It allows the estimation of the overlap between the corresponding classes, which is a measure of correlation between them.
Arguments about whether a robot could ever be conscious have been conducted up to now in the factually impoverished arena of what is `possible in principle'. A team at MIT, of which I am a part, is now embarking on a longterm project to design and build a humanoid robot, Cog, whose cognitive talents will include speech, eye-coordinated manipulation...
Neural state machines are known to have universal computing
powers, which, through methods such as iconic learning can be turned
towards cognitive tasks. This makes the division of allegiances to
symbols or neurons appear somewhat false, not to say counterproductive.
The states of the net, the states of the input and output are clearly
symbols that...
A neural function is developed that combines the characteristics
of weightless and weighted binary neurons. A new combined generalisation
algorithm is presented and applied to a neural state machine which is
capable of learning to respond to sequences of inputs. The difficulty
with such tasks lies in learning appropriate internal state assignments....
This paper is about communication between people and machines and the stored knowledge that a machine must possess to “understand” the transaction. Neural state machines are proposed which build up their knowledge iconically and represent both visual and linguistic knowledge
There has been a recent renewal of interest in the debate as to whether an explanation of consciousness can (Dennett, 1991) or cannot (penrose, 1989) be captured by some formal theory. The aim of this paper is to side with the former and to present a theory testable through the concept of a neural state machine. This is a development of a theme fir...
This paper presents a solution to the problem of determining the
distribution of the n-dimensional Boolean space in relation to two
arbitrary patterns which are separated by a fixed distance h in such
space. The estimation of such distribution allows the evaluation of the
amount of overlap between the two corresponding r-spheres, which is a
measure...
Nature is the international weekly journal of science: a magazine style journal that publishes full-length research papers in all disciplines of science, as well as News and Views, reviews, news, features, commentaries, web focuses and more, covering all branches of science and how science impacts upon all aspects of society and life.
Three key points are made in this paper. The first is that iconic states arise naturally in all recursive neural nets which are trained on world states. The second point is that this leads to a new paradigm of cognitive representation. Language is seen as a vehicle for retrieving iconic representations in a recursive system — this is dubbed the ico...
"Neurons and Symbols", the successor to the authors' work "An Introduction to Neural Computing", presents in a unified explanatory style the emerging points of a fierce contemporary debate on the effect that neural models are likely to have on more classic symbolic models of cognition. The controversy is multi-disciplinary - it involves scientists...
The key element presented in this paper is an adaptive state machine in which neurons form the state variables and these are sufficient in number for the state to be “iconic”. That is, the state reflects events in the world in a many-to-some manner. Early processing forms part of this model. The model is a variant of the previously defined Neural S...
In a totally reduced view of the interaction of a learning machine with a “world”, such a machine can receive world-state information, it can act on the world and it can receive data from the world that controls learning. We see the learning machine as a classical state machine whose states are its constructs of the experienced world. Learning is a...
A full rationale is presented for designing systems with enhanced cognitive properties using RAM-based or “weightless” methods. The paper concentrates on a structure with associative properties called the General Neural Unit (GNU) and shows that weightless analysis leads to an understanding of storage capacity and retrievability of stored training...
Nature is the international weekly journal of science: a magazine style journal that publishes full-length research papers in all disciplines of science, as well as News and Views, reviews, news, features, commentaries, web focuses and more, covering all branches of science and how science impacts upon all aspects of society and life.
Nature is the international weekly journal of science: a magazine style journal that publishes full-length research papers in all disciplines of science, as well as News and Views, reviews, news, features, commentaries, web focuses and more, covering all branches of science and how science impacts upon all aspects of society and life.
The revival of interest in neural networks is now not only well known, but is a discipline which is being embraced by many. This paper is a brief cautionary tale of the long way that still lies ahead before neural networks become as well understood as other structures that come within the domain of the electrical engineer