Blay Whitby - Artificial Intelligence

Artificial Intelligence

A Beginners Guide

ONEWORLD BEGINNERS GUIDES combine an original, inventive, and engaging approach with expert analysis on subjects ranging from art and history to religion and politics, and everything in between. Innovative and affordable, books in the series are perfect for anyone curious about the way the world works and the big ideas of our time.

aesthetics

africa

anarchism

aquinas

art

artificial intelligence

the bahai faith

the beat generation

biodiversity

bioterror & biowarfare

the brain

british politics

the buddha

cancer

censorship

christianity

civil liberties

classical music

climate change

cloning

cold war

conservation

crimes against humanity

criminal psychology

critical thinking

daoism

democracy

descartes

dyslexia

energy

engineering

the enlightenment

epistemology

evolution

evolutionary psychology

existentialism

fair trade

feminism

forensic science

french literature

french revolution

genetics

global terrorism

hinduism

history of science

humanism

huxley

islamic philosophy

journalism

judaism

lacan

life in the universe

literary theory

machiavelli

mafia & organized crime

magic

marx

medieval philosophy

middle east

NATO

nietzsche

the northern ireland conflict

oil

opera

the palestineisraeli conflict

paul

philosophy of mind

philosophy of religion

philosophy of science

planet earth

postmodernism

psychology

quantum physics

the quran

racism

renaissance art

shakespeare

the small arms trade

the torah

sufism

volcanoes

A Oneworld Book
Published by Oneworld Publications 2003
This ebook edition published 2012

Copyright Blay Whitby 2003

The moral right of Blay Whitby to be identified as the Author of this work has been asserted by him in accordance with the Copyright, Designs and Patents Act 1988

All rights reserved
Copyright under Berne Convention
A CIP record for this title is available from the British Library

ebook ISBN 9781780741291
ISBN 9781851686070

Typeset by Jayvee,Trivandrum, India
Cover design by Two Associates

Oneworld Publications
185 Banbury Road
Oxford
OX2 7AR
England

Stay up to date with the latest books, special offers, and exclusive content from Oneworld with our monthly newsletter

Sign up on our website

www.oneworld-publications.com

For Sharon Wood for introducing me to the field of Artificial Intelligence

Contents

Preface

This book is intended to introduce the reader to the fascinating world of Artificial Intelligence. Like the other titles in the Oneworld series, it presumes no previous knowledge of the area but assumes that both the subject and the excitement around it can be conveyed to a general audience in a non-technical and jargon-free manner.

Artificial Intelligence (normally abbreviated to AI) is one of the most possibly the most exciting challenges that humanity has ever undertaken, although it is hard perhaps to convey excitement within the pages of a relatively serious book. However, the unrestrained gusto with which researchers from the widest possible variety of backgrounds throw themselves into the many diverse challenges of AI makes it a very exciting place to work. Some of the challenges described may, at first glance, seem overambitious but reaching for the stars and ignoring the cynics is characteristic of a young science. I have been writing on AI for twenty years and have no illusions as to just how difficult some of its goals are to attain. However, I consider it an inestimable privilege to have been in the same room as people who, on being told that something was impossible, promptly went out and did it.

AI is not only all about ambitious remote goals: it is also a remarkably successful technology. Much of what was once presented as cutting-edge ideas is now regarded as everyday computing technology. Ex-students and colleagues who have entered the commercial sector have, on occasion, made small fortunes from developing and exploiting AI technologies.

There is no shortage of good books on AI. However, few of them are particularly suited to introduce a beginner to the field. Student texts have to deal with the subject in a more technical manner and fluency with various types of formal notation is usually required.

AI is an extremely diverse area and it would be impossible for this book (or possibly any book) to cover every important piece of research and technology in detail. In order to introduce the general reader to as much of the area as possible I have inevitably had to be selective and to describe some of the technical details rather briefly. The bibliography should allow the curious reader to fill in any missing details.

Because this is a book for the general reader I have refrained from introducing any equations, computer code, logic, or symbolic descriptions. I have also kept scholarly references to a minimal level. In this sense, it is very different from an AI textbook of which there are many excellent examples available. My personal favourites are listed in the further reading sections for the relevant chapters (including this preface) but omission of a title does not imply any criticism.

This book alone, therefore, cannot prepare you for a job in AI. If you are interested in pursuing a career in AI you should, if at all possible, learn some computer programming. This is not because AI researchers spend all their time writing computer programs in fact they tend to spend most of their time thinking about difficult problems it is simply because computer programming is a fundamental skill in this area and the closest thing there is to a lingua franca in AI.

Two myths about AI

There are two particularly pernicious myths about AI which are in general circulation and which this book will comprehensively dismiss. The first myth is the claim that AI has failed. (Some people occasionally make the stronger claim that AI is impossible.) This myth is just plain false. AI has succeeded in producing a steady stream of successful technologies that are changing the way we live. Examples of the successes of AI litter the chapters of this book. There can be little doubt that the activities of AI researchers will continue to spin off useful everyday technologies. AI has also succeeded in producing a large number of important ideas and methods that have spectacularly influenced other areas of science and art.

Why such a myth might persist probably lies in a misperception of the goals of AI. It is easy to become confused about these goals since AI has and always has had multiple goals ranging from making computers easier to use, to understanding human thought. There have always been those who expressed cynicism about AI both inside and outside the field. Fluidity of goals and methods is typical of, and probably helpful to, an area of science that is still very young.

The second myth is the notion that AI, when it succeeds, will leave humanity in a state of serfdom to a new race of all-powerful machines. This second myth is not merely false: it is also nonsensical. It must be admitted that some of the responsibility for the introduction of this myth lies with certain individuals within AI. Even so, it makes no sense. There is nothing in present-day or foreseeable AI that is likely to bring about this dystopian prospect. AI has, so far, proved to be a remarkably benign technology. There is, as with all technologies, a balance of costs and benefits which we will consider in detail. However, AI compares very favourably with many other technologies of the second half of the twentieth century in this respect.