Artificial Consciousness

Antonio Chella

Description

The book is interdisciplinary and focuses on the topic of artificial consciousness: from neuroscience to artificial intelligence, from bioengineering to robotics. It provides an overview on the current state of the art of research in the field of artificial consciousness and includes extended and revised versions of the papers presented at the International Workshop on 'Artificial Consciousness', held in November 2005 at Agrigento (Italy).

Sujets

• Sociologie, société et politique
• Artificial intelligence

Informations

• Publié par: Andrews UK
• Date de parution: 01 octobre 2013
• Nombre de lectures: 0
• EAN13: 9781845406776
• Langue: English
• Poids de l'ouvrage: 1 Mo

Extrait

Title Page
ARTIFICIAL CONSCIOUSNESS
Edited by
Antonio Chella
and
Riccardo Manzotti



Publisher Information
Originally published in the UK by
Imprint Academic
PO Box 200
Exeter
EX5 5YX, UK
Originally published in the USA by
Imprint Academic
Philosophy Documentation Center
PO Box 7147
Charlottesville
VA 22906-7147, USA
www. imprint-academic.com
Digital edition converted and distributed in 2013 by
Andrews UK Limited
www.andrewsuk.com
Copyright © Antonio Chella and Riccardo Manzotti, 2007
The moral rights of the contributors have been asserted. No part of any contribution may be reproduced in any form without permission, except for the quotation of brief passages in criticism and discussion.



Foreword
Do Engineers Dream of Conscious Robots?
Quam multa fieri non posse, priusquam sint facta, judicantur!
(How many things, that were held to be impossible, have been accomplished!)
Plinius the Elder, 70 CE.
A slight change in the title of Phil K. Dick’s famous novel might help us in understanding the dream pursued by some engineers. Our answer to such a question is yes. As far as we know, natural selection developed conscious agents - human beings for sure, almost certainly most mammals, perhaps other species. Engineers have always tried to mimic nature: helicopters and aeroplanes for mimicking birds, steam engines for challenging the power of horses, computers for imitating the logical capability of humans, submarines to ape fishes. Can engineers succeed in replicating what seems to be one of the most intimate and private features of human beings, namely consciousness?
At the beginning of the twenty-first century, some engineers are dreaming of conscious robots, not just intelligent robots. For the first time in the history of engineering, a term so vague as consciousness is taken into consideration in the technical world (Steels, 1995; Schlagel, 1999; Jennings, 2000; Aleksander, 2001; Buttazzo, 2001; Holland, 2003). What do they mean exactly by ‘conscious robots’, ‘artificial consciousness’ or ‘machine consciousness’? This book [1] gathers state of the art work of researchers engaged worldwide in the seemingly oxymoronic field of artificial consciousness. It follows up the previous effort made by Owen Holland four years ago (Holland, 2003). In English we have two related terms: consciousness and conscience. The former refers to phenomenal experience. The latter refers to ethical and moral issues. Engineers don’t talk about conscience: they talk about consciousness. Some authors held that if nature has been capable of producing conscious beings, through natural selection, there is no reason why the same should not be done in artefacts, through design and craft.
The nature and mechanism of consciousness are still far from being understood. There is no consensus on what consciousness is and how it must be studied. This situation is at the same time an opportunity and a curse for the engineers engaged in such an endeavour. It is an opportunity because very often, in the history of engineering, engineers built artefacts even if they were unaware of the underlying scientific principles. It is a curse because engineers prefer to apply formalized and well-established techniques.
The former position is well expressed by Gerard Edelman and Giulio Tononi’s words (Edelman and Tononi, 2000):
To understand the mental we may have to invent further ways of looking at brains. We may even have to synthesize artefacts resembling brains connected to bodily functions in order fully to understand those processes. Although the day when we shall be able to create such conscious artefacts is far off we may have to make them before we deeply understand the processes of thought itself.
The latter position is succinctly expressed by Ron Arkin (Arkin, 1999), p. 427: ‘Most roboticists are more than happy to leave these debates on consciousness to those with more philosophical leanings.’ All those gathered in the present book think otherwise.
Although we are aware that there is no commonly accepted crystal clear definition of consciousness, we do not think that this lack ought to hinder the scientific research of consciousness. We agree with the standpoint adopted by Christof Koch (Koch, 2004), p. 12:
Historically, significant scientific progress has commonly been achieved in the absence of formal definitions. For instance, the phenomenological laws of electrical current flow were formulated by Ohm , Ampère, and Volta well before the discovery of the electron in 1892 by Thompson. For the time being, therefore, I adopt a working definition of consciousness and will see how far I can get with it.
Engineers are trained in a reductionist cultural environment. Consciousness seems unfit to such reductionist standpoint. Nevertheless consciousness beings are a fact. Current science developed an objectivistic ontology of nature that seems incapable of dealing with subjectivity. The recent upsurge of interest in the scientific study of consciousness has raised methodological and empirical problems. At the intersection between engineering and current scientific approaches to consciousness lies the seminal field of artificial consciousness. As Igor Aleksander puts the matter (Aleksander, 2000): ‘This was the moment when the seemingly incongruous idea of an artificial consciousness came into being. Did the concept make sense?’
Artificial consciousness is apparently an oxymoron since the artificial is the domain of human artefacts, thus of objects, and the quintessential feature of consciousness is subjectivity. Subjects and objects are an uneasy match. Designing and implementing an artificial conscious machine appears an impossible task. However, it is possible that in the course of attempting it we shall develop new concepts and reformulate old categories.
It is worthwhile to stress that, since mid-eighties, in the field of engineering, the issue of consciousness has come out in many different contexts. The crisis of artificial intelligence, the new perception of artificial physical bodies, the relevant progress in brain sciences have shifted the attention of many researchers from the study of intelligent behaviour to the study of consciousness: how agents develop the capability of having experience, of being aware of what happens to them and around them.
The distinction between subjects and objects once seemed clear. Subjects have the capability of having experiences, of being aware of what happens to them and around them - that is, they are conscious. Objects are unconscious. But things are changing. As the novelist David Lodge wrote in one of his tales (Lodge, 2001), p. 36: ‘Once upon a time nobody was interested in the problem of consciousness except a few philosophers. Now it’s the biggest game in town.’ In this cultural framework, engineers try to participate. Inside this group, Igor Aleksander was the first one in 1992 to mention the term ‘artificial consciousness’ in a scientific paper (Aleksander, 1992).
Even if the boundaries, methods and goals of this field are still rather unclear, even if we have not yet outlined measurement methods for the appearance of consciousness in artefacts, many researchers are trying to develop models of consciousness that can be implemented in machines.
Nevertheless, it seems that there is some kind of ontological mistake that thwarts any attempt to deal explicitly with consciousness. The aim of this book is to understand why it is so difficult to approach the problem of subjectivity from an engineering viewpoint and, then, to propose alternative frameworks that could cope with the design and implementation of a conscious subject. Any proposal for ontological revision must not remain a sterile metaphysical project but must be tested empirically. That is where engineers might provide a unique opportunity to advance our understanding of consciousness. Robotics is a field in which experiments might be carried out. If there were a theory of consciousness, which sets the conditions by which an object could let a subject emerge, such conditions could be replicated. Hitherto, there have been only a few sparse attempts to understand and propose an architecture capable of producing a conscious machine.
The book collects many different authors whose views and methods are rather different. A rule of thumb to compare them is based on their opinion about the nature of subjectivity. For a few of them subjectivity must be reduced to objective or functional processes. For a few others, subjectivity will require a conceptual change in our fundamental categories.
The book is organized in three parts: the race for artificial consciousness, design and implementation of an artificial machine, and artificial and natural consciousness.
In the first part the authors introduce the reader to the current situation of the field, highlighting the relation between artificial consciousness and previous approaches like artificial intelligence, cognitive science, neuroscience, and philosophy of mind.
Vincenzo Tagliasco introduces the reader to the undoubted roots of artificial consciousness, tracing its origins in first years of cybernetics and artificial intelligence. Sketching a brief history of the theme of ‘artificial consciousness’ in the scientific and technological panorama, he explores the relation between different cultural traditions in dealing with the issue of an artificial subject. Finally, he highlights the implicit legacy between this new field and other areas such as cinema and literature.
In the second chapter, John Taylor analyzes the principles on which machine consciousness should be based, as arising from crucial features of human consciousness. Employing present kn