Artificial Beings E-Book

Jacques Pitrat

• Herausgeber: John Wiley & Sons
• Kategorie: Wissenschaft und neue Technologien
• Sprache: Englisch

Seitenzahl: 544

Veröffentlichungsjahr: 2013

Beschreibung

It is almost universally agreed that consciousness and possession of a conscience are essential characteristics of human intelligence. While some believe it to be impossible to create artificial beings possessing these traits, and conclude that ultimate major goal of Artificial Intelligence is hopeless, this book demonstrates that not only is it possible to create entities with capabilities in both areas, but that they demonstrate them in ways different from our own, thereby showing a new kind of consciousness. This latter characteristic affords such entities performance beyond the reach of humans, not for lack of intelligence, but because human intelligence depends on networks of neurons which impose processing restrictions which do not apply to computers.

At the beginning of the investigation of the creation of an artificial being, the main goal was not to study the possibility of whether a conscious machine would possess a conscience. However, experimental data indicate that many characteristics implemented to improve efficiency in such systems are linked to these capacities. This implies that when they are present it is because they are essential to the desired performance improvement. Moreover, since the goal is not to imitate human behavior, some of these structural characteristics are different from those displayed by the neurons of the human brain - suggesting that we are at the threshold of a new scientific field, artificial cognition, which formalizes methods for giving cognitive capabilities to artificial entities through the full use of the computational power of machines.

Leseprobe

Contents

Acknowledgements

Note on the Terminology

Chapter 1. Presenting the Actors

1.1. The book

1.2. Human and artificial beings

1.3. The computer

1.4. The author

1.5. CAIA, an artificial AI scientist

1.6. The research domains of CAIA

1.7. Further reading

Chapter 2. Consciousness and Conscience

2.1. Several meanings of “consciousness”

2.2. Extending the meaning of “conscience” for artificial beings

2.3. Why is it useful to build conscious artificial beings with aconscience?

2.4. Towards an artificial cognition

Chapter 3. What Does “Itself” Mean for an Artificial Being?

3.1. Variousversions of an individual

3.2. Variants of an individual

3.3. Cloningartificial beings

3.4. Dr. Jekyll and Mr. Hyde

3.5. The Societyof Mind

3.6. More onthe subject

Chapter 4. Some Aspects of Consciousness

4.1. Six aspects of consciousness

4.2. Some limits of consciousness

Chapter 5. Why is Auto-observation Useful?

5.1. Auto-observation while carryingout a task

5.2. Auto-observation after the completion of a task

Chapter 6. How to Observe Oneself

6.1. Interpreting

6.2. Adding supplementary orders

6.3. Using timed interruptions

6.4. Using the interruptionsmade by the operating system

6.5. Knowing its own state

6.6. Examining its ownknowledge

6.7. The agents of the Societyof Mind

6.8. The attention

6.9. What is “I”

Chapter 7. The Conscience

7.1. The conscience of human beings

7.2. The conscience of an artificial being

7.3. Laws forartificial beings

Chapter 8. Implementing a Conscience

8.1. Why is aconscience helpful?

8.2. The conscience of CAIA

8.3. Implicit principles

8.4. Explicitprinciples

8.5. The consciences in a societyof individuals

Chapter 9. Around the Conscience

9.1. Emotions

9.2. Changing its conscience

9.3. A new human conscience for our relationships with artificial beings

Chapter 10. What is the Future for CAIA?

Appendices

1. Constraint Satisfaction Problems

2. How to implement some aspects of consciousness

Bibliography

Index

First published in Great Britain and the United States in 2009 by ISTE Ltd and John Wiley & Sons, Inc.

Apart from any fair dealing for the purposes of research or private study, or criticism or review, as permitted under the Copyright, Designs and Patents Act 1988, this publication may only be reproduced, stored or transmitted, in any form or by any means, with the prior permission in writing of the publishers, or in the case of reprographic reproduction in accordance with the terms and licenses issued by the CLA. Enquiries concerning reproduction outside these terms should be sent to the publishers at the undermentioned address:

ISTE Ltd27-37 St George's RoadLondon SW19 4EUUK

John Wiley & Sons, Inc.111 River StreetHoboken, NJ 07030USA

www.iste.co.uk

www.wiley.com

© ISTE Ltd, 2009

The rights of Jacques Pitrat to be identified as the author of this work have been asserted by him in accordance with the Copyright, Designs and Patents Act 1988.

Library of Congress Cataloging-in-Publication Data

Pitrat, J. (Jacques) Artificial beings : the conscience of a conscious machine / Jacques Pitrat.

p. cm.

Includes bibliographical references and index. ISBN 978-1-84821-101-8

1. Artificial intelligence--Moral and ethical aspects. 2. Conscience. 3. Consciousness. I. Title.

Q335.P558 2009

006.3--dc22

2009005292

British Library Cataloguing-in-Publication Data

A CIP record for this book is available from the British Library

ISBN: 978-1-84821-101-8

Acknowledgements

First and foremost, I wish to thank Basile Starynkevitch and Carla who carefully read the first versions of this book, and suggested many improvements.

I also want to extend my gratitude to Robin who has been the system administrator of my computer since I retired from the CNRS. He has managed it so well that everything has run smoothly during all these years.

Work on this system began while I was a CNRS researcher until my retirement in 2000. I worked at the LIP6, a laboratory of Pierre and Marie Curie University (Paris VI), and I wish to thank all the members of the laboratory who assisted me in various ways during the first years of the development of my system. I am forever grateful to the members of the Métaconnaissances research team who provided intelligent comments and useful criticisms to my work.

I am grateful to the editors and readers of ISTE Ltd whose advice and suggestions were most helpful.

Note on the Terminology

Artificial beings (in contrast to human beings) are the subject of this book. As up to now most of their activities have been performed by humans, an extensive vocabulary can be used to describe actions when the performer is human, but few words are available for when it is an artificial being. Therefore, I will speak of artificial beings which think, choose, prefer, are conscious, want, etc. However, when I say that an artificial being “thinks” that only means that it processes some data so that it generates new data, which will help it to determine what it will do. For instance, it gives up its present task because it is too difficult, it creates a method for solving problems, it chooses the next rule that it will execute, it tries to understand the reason for a previous mistake, etc.

There is no anthropomorphism intended in this use of the vocabulary. It is simply due to the fact that it is impossible to create a new word for each activity of an artificial being when it is in the same situation as a human being, and when its results are often as good as ours. This does not mean that I believe that an artificial being operates as a human does. On the contrary, I will try to show that they can process information in a completely different way, particularly when they observe their behavior, which gives them a large advantage over us.

Let us consider an example: when I speak of the artificial being which I built, I say that it is autonomous. When we say that a human being is autonomous, that means that he can make his decisions without asking for help. In this meaning, my creature is certainly autonomous, it works without any interference for more than one month. However, this characteristic is not enough; a traffic light can work alone for years, it does not need an exterior agent to go from green to orange or from orange to red; an autonomous agent must not be completely predictable. However, this point raises a problem, because artificial beings are usually completely determined, they are following a program and, theoretically, we can predict all of their future decisions when we know their program and their data. Naturally, this comparison is unfair because we do not know how our brain works, even if it is as determined as a computer program. We cannot reject the possibility of an autonomous artificial being only because we do not know how human beings work. Moreover, theoretically predictable does not mean practically predictable. It is impossible to predict the behavior of a complex artificial being, it depends on its history and on a huge number of small operations, we will never have enough time to execute them. At different instants, it can take different decisions although it is in exactly the same situation because they also depend on the events that happened between these instants. The behavior of an artificial being may also be more difficult to predict than the behavior of a human one. For these reasons, I will speak of autonomous artificial beings, although it does not choose its act in the same way as us.

Finally, in Artificial Intelligence (AI) it is important to indicate unambiguously whether we are considering a human or an artificial being, particularly when we use pronouns. Indeed, both kinds of beings have similar activities and the context does not always indicate the nature of the agent. Thus, I will always use “it” to refer to an artificial being, which will often be my system CAIA. When I consider an activity in common with human and artificial beings, the agent is often represented by “one” as in “Let us assume that one has decided to interpret a program”; all we know of this being is that is able to make a decision, and it is not specified whether this agent is human or artificial. The passive form is also convenient to mention a being without specifying its nature of an actor as in “when the program has been written”. In some situations, to insist on the fact that I am simultaneously speaking of human and artificial beings, I will write his/its, he/it, or who/which. I will never use his/her/its, the situation is complicated enough with two kinds of actors: when I introduce a human being, he may be male or female in all this book, even when I use “man” or “he”. “We” will always represent human beings, the reader–author pair as in “We will later see that ...” or all human beings as in “We do not like the idea that artificial beings could become more intelligent than human beings”.

Chapter 1

Presenting the Actors

At the beginning of a book, it is natural to present its actors. Of course, the first actor is the book itself. Then I will examine the qualities and drawbacks of two important families of actors: human beings and artificial beings. I do not forget the computer, an essential support for an artificial being. Next, I will present my reasons for developing my present research, and introduce the subject of this research: my colleague CAIA, which is an artificial scientist. Studying its behavior and its mechanisms will supply most of the examples of this book. Finally, I look at the domains where CAIA is carrying out research.

1.1. The book

The goal of this book is to examine whether an artificial being can have some capacities similar to those that consciousness or conscience give to human beings. Over many centuries, a huge amount of work has been done on human consciousness and conscience, but sometimes I feel that I still do not understand them. In relation with my work, to understand a hypothesis on the working of our brain, I need to see how the hypothesis can be implemented in a computer program. Naturally, scientists who lived before the computer era could not express themselves in such a precise way because they did not know of computers and how we can program them. This is a difficult and unnatural way of thinking, and even now, we have to realize several Artificial Intelligence systems before we can easily feed computer programs with attractive ideas. When I read these early books, I can see a dozen ways to implement each of their ideas and so I feel confused: among all these possibilities, which one was seen by the author? For instance, I have never understood how the qualia could be exactly represented. The qualia are at the core of many theories of consciousness, but philosophers agree neither on their meaning, nor on their properties. We can roughly say that their goal is to represent in our brain sensations like redness, but when the authors want to be more precise, each one has his own definition. How could we use this concept in a computer program?

Over several years, some outstanding books and papers, such as those of Marvin Minsky, Daniel Dennett, Grard Sabah or Franois Anceau, have been written by scientists who know exactly how a computer works, and the reader can understand how their ideas might be implemented. However, these ideas have not always been effectively implemented. An AI scientist knows that it is impossible to foresee in a program all the elements that would be necessary to give excellent results: either some mechanisms are forgotten, or their description is not sufficiently accurate. We have to experiment with the system, and modify it to work better. Thus, a description without an implementation is an indispensable starting point, but it is not sufficient, although it contains many interesting and promising ideas. Moreover, the goal of most of the research has been to improve our understanding of consciousness and of conscience for human beings. Usually, they were not concerned to see if these faculties could be expressed by artificial beings, which have different ways to process information. However, there are some exceptions; in particular, John McCarthy has written a theoretical paper on the possibility of making robots conscious of their mental states. My book complements his work and I do not consider the theoretical point of view, but rather its realization by a practical computer system. His paper must therefore be read by all those who are interested in this approach. Several papers have been published in the last years on the realization of various aspects of consciousness. They contribute new and interesting ideas, but this is not enough: intelligence is a whole, and we cannot examine only how some of its aspects can be implemented, we have also to study what kind of consciousness has been given to artificial beings which have to perform difficult tasks.