Review - The Cognitive Approach to Conscious Machines

The author in this book argues that the conscious machines can be built, but rejects artificial intelligence and classical neural networks in favor of the emulation of the cognitive processes of the brain, proposing a special cognitive architecture to reproduce the perceptual processes, inner imagery, inner speech, pleasure, pain, emotion, and the cognitive functions behind these. He proposes a realization of novel artificial neurons and non–numeric signal-level representations that carry dedicated meanings. The machine is expected to produce higher-level functions by the power of the elementary processing unit, the artificial neuron, without algorithms or programs. The dedicated chips would have to be designed, consequently for such a machine.

"The real challenge of consciousness is its apparent immaterial nature." The author raises questions like, "Does a conscious machine really perceive the flow of inner speech and imagery as immaterial?", "Does it perceive the possession of an immaterial self?", and "Is it aware of its own existence?"(5) The book presents author's answer- proposals to these and other similarly related questions.

The book, to begin with, presents a brief history and a review of computing principles, arguments against the concept of thinking as the execution of strings of program commands, artificial neural networks as a better way to realize such a machine.

The book is in three Parts. Part 1 is on thinking and computation and on artificial neural networks considered as to the latter's rescue. Part II is on cognition and consciousness and covers all the processes referred to in the first paragraph above, and Part III presents design philosophy and model for machine cognition and consciousness, discussing the general nature of consciousness in the light of this model.

Part III is critical to the proposed theme of the book. The author intends developing and proposing machine models that could emulate or duplicate the processes of human cognition and even consciousness. To what extent can this be done? "Could there be any convincing arguments that a cognitive machine would possess any of the sought after properties? What if anything would the design of cognitive machines tell about the problem of consciousness? What should this machine be like and how would it differ from existing computers?" (169).

What has followed in Part III, now, as a sequel to author's commitment to the questions above are a series of nine chapters which present discussion on: Machine Models for Cognition and Consciousness-Introduction, Representation of Information, Artificial Associative Neurons and Neuron Groups, Models for Machine Perception Processes, A Cognition Machine Architecture, Models for Motor Functions, Machine Emotions, Language and Inner Speech, Inner Imagery and Thinking, Machine Consciousness, and Technology and Final Questions. Arising out of these discussions and considering several other aspects of the modeling process, such as "A complicated environment, conflicting lists of tasks, goals and needs as well as evoked memories and possibilities for inference would be a real challenge to any cognitive system.... .(168)", the author outlines the general requirements for a cognitive machine. The requirements proposed are:(1) "A suitable method for the representation of information must be devised. (2) Suitable information processing elements that allow the manipulation of information by the chosen representation method must be designed. (3) A machine architecture that can accommodate censors, effectors, the processes of perception, introspection and the grounding of meaning as well as the flow of inner speech and inner imagery must be designed. (4) The system design must also accommodate the functions of thinking and reasoning, emotions and language (168)."

According to the author, his approach is not entirely theoretical. He has developed experimental chips that are suitable for the implementation of proposed cognitive architecture. There are others also who have achieved similar goals. "New technology will arise," to change the future. However, according to him, as also according to others, "Will machine consciousness, implemented on a chip, resolve the philosophical questions about mind and consciousness for once and all" (263)? He admits, however though, that "many questions will be clarified and become trivial as we become familiar with the consciousness of the machine" (263), none the less, many questions still remain open.

One continues to ponder about Skillful Robots, Mind on Silicon, and Consciousness and Self or Raise the Dead as realizable enigma. He talks about Star Trek Movie and Cloning, and their realizability and the conceptual and logical questions that come up in this context.

In this concluding section, the author lets loose his imagination, discussing immediately realizable and remote-future possibilities. Virtual minds," the author ponders, are the copies, more or less accurate of the originals yet the thin consciousness of these would not be that of the original. "Would there be a way to capture the real consciousness and identity of a dead person and could we transfer the consciousness of a living person into a machine?"

To what extent cloning can be used to this advantage employing the "technique of implanting artificial memories, known to psychology nowadays. Given enough time, all essential personal history might be implanted by employing means that are available now. The authors continues to stretch his imagination to examine the possibility of `raising the dead.' However, "a perfect biological copy, even with a perfectly copied mind, is not the original" (268).

The author, while concluding, raises other issues such as "Getting into the Machine," "Final Question: Being Inside," "Our Existence", "Identity, and scope of our Consciousness." He concludes by stating whether a "conscious machine also, due to the apparent immaterial nature of its thoughts, reproduce Descartes' error and infer the existence of immaterial mind? Will the machine, too, come to the conclusion that there is something special in being here, to exist as a self-conscious individual"?

During the course of discussions the author raises several other issues as for instance that of the Easy and Hard Problem of Consciousness, described by the following terms: qualia or feel of the percepts, phenomenal consciousness, conscious experience, what does it mean when a conscious subject is phenomenally feeling something. "Will not the conscious machine operate without any felt sensations in a kind of zombie like manner? Would this kind of system, if complex enough, appear to have consciousness? May be, and it might even insist that it has subjective experiences" (146).

It is a fascinating presentation of various issues relating to the emulation of consciousness by a machine and its capabilities. The problems that confound the issues have been so clearly and precisely presented that even a beginner student will have no difficulty in getting at the meanings. For the undergraduates it will be a comprehensibly readable text. Meaningfully precise presentation has gone a long way to support the quality of text.

Dr. G.C. Gupta, Visiting Professor of Cognitive Science, University of Allahabad, Allahabad, India

Share:

Link: (Here's how to do it ... )