Societal Responsibility of Artificial Intelligence E-Book

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

Seitenzahl: 381

Veröffentlichungsjahr: 2021

Beschreibung

The digital world is characterized by its immediacy, its density of information and its omnipresence, in contrast to the concrete world. Significant changes will occur in our society as AI becomes integrated into many aspects of our lives.

This book focuses on this vision of universalization by dealing with the development and framework of AI applicable to all. It develops a moral framework based on a neo-Darwinian approach - the concept of Ethics by Evolution - to accompany AI by observing a certain number of requirements, recommendations and rules at each stage of design, implementation and use. The societal responsibility of artificial intelligence is an essential step towards ethical, eco-responsible and trustworthy AI, aiming to protect and serve people and the common good in a beneficial way.

Leseprobe

Table of Contents

Cover

Title Page

Copyright

Acknowledgments

Introduction

1 Societal and Moral Questioning Around AI and Its Ecosystem

1.1. Use cases of AI

1.2. Digital environment

1.3. What is the place for human beings in this digital society?

1.4. Technological and societal issues

1.5. Ethical and moral issues

2 The Ethical Approach to AI

2.1. Definition of ethics

2.2. General ethical principles

2.3. Problems and ethical issues specific to the digital environment

2.4. Ethical criteria and better risk assessment of AI-related digital projects

2.5. Analysis of AI-related knowledge

3 Ethical Framework Associated with AI

3.1. Ethical charter around AI

3.2. Recommendations for AI

3.3. Temporality relative to the human guarantee in digital technology

3.4. For the health user and for health user representation

3.5. For health personnel and for the representation of health personnel

3.6. Environmental parameters of digital technology

3.7. Regulation associated with AI

3.8. Algorithmic systems and digital data governance

3.9. Four key steps for an AI project

3.10. Algorithmic responsibility

4 Anticipation Around Artificial Consciousness

4.1. Protean aspects of consciousness associated with intelligence

4.2. Structuring of consciousness

4.3. Neoplatonic systemic ethical modeling (Ψ, G, Φ) of an artificial consciousness

4.4. Process of creating practical wisdom from artificial consciousness

4.5. Morality of a “strong” AI

Conclusion

Appendices

Appendix 1 Ethical Charter of Using AI in Judicial Systems and Their Environment

Appendix 2 Practical Recommendations of the CNIL Regarding the Ethics of Algorithms

Appendix 3 OECD Recommendation on AI

Appendix 4 Questions Concerning the Application of Ethical Standards

A4.1. Question 1: universality of standards

A4.2. Question 2: moral saturation

A4.3. Question 3: bias

A4.4. Question 4: integration and compatibility of standards

A4.5. Question 5: trust

A4.6. Question 6: environment

Appendix 5 CERNA Recommendations on Machine Learning

Appendix 6 Reasons for a “Digital Divide”

Appendix 7 Holberton–Turing Oath

A7.1. Holberton–Turing oath

Appendix 8 Report Proposals: “For a Controlled, Useful and Demystified Artificial Intelligence”

A8.1. For a controlled artificial intelligence

A8.2. For useful AI, in the service of humans and humanistic values

A8.3. For a demystified AI

List of Abbreviations

References

Index

End User License Agreement

List of Illustrations

Chapter 2

Figure 2.1. Ethical categories associated with algorithmic processing. For a col...

Figure 2.2.

Ethical criteria for algorithmic processing

Figure 2.3. Study of the knowledge pyramid through ethical modeling. For a color...

Chapter 4

Figure 4.1. Modeling an artificial consciousness through the Neoplatonic systemi...

Figure 4.2. Process of creating practical wisdom from an artificial consciousnes...

List of Tables

Chapter 1

Table 1.1.

AI use cases by industry

Table 1.2.

Digital ecosystem of Big Data operated by AI

Chapter 2

Table 2.1.

Vocations and questions of ethical principles applied to AI

Table 2.2.

Impacts of AI data on individuals

Table 2.3.

Positive and negative effects of AI in the socioecological sector

Table 2.4.

Structuring of ethics of algorithms

Table 2.5.

Parameters constituting the environment surrounding knowledge

Table 2.6.

Structuring of the knowledge pyramid

Chapter 3

Table 3.1. Steps for integrating an approach to Ethics by Evolution within a dig...

Table 3.2. Temporality associated with the human guarantee in the case of AI in ...

Table 3.3.

Strategic launching of AI within a structure

Table 3.4.

Questions and actions to prepare an AI-related project

Table 3.5.

SWOT table of AI issues for a company

Chapter 4

Table 4.1.

Links between levels of consciousness and forms of intelligence

Table 4.2. Interactions between levels of ethics, dimensions of systems modeling...

Guide

Cover

Table of Contents

Title Page

Copyright

Acknowledgments

Introduction

Begin Reading

Conclusion

Appendices

Appendix 1 Ethical Charter of Using AI in Judicial Systems and Their Environment

Appendix 2 Practical Recommendations of the CNIL Regarding the Ethics of Algorithms

Appendix 3 OECD Recommendation on AI

Appendix 4 Questions Concerning the Application of Ethical Standards

Appendix 5 CERNA Recommendations on Machine Learning

Appendix 6 Reasons for a “Digital Divide”

Appendix 7 Holberton–Turing Oath

Appendix 8 Report Proposals: “For a Controlled, Useful and Demystified Artificial Intelligence”

List of Abbreviations

References

Index

Other titles from ISTE in Health Engineering and Society

End User License Agreement

Pages

v

iii

iv

ix

xi

xii

xiii

xiv

xv

xvi

xvii

xviii

xix

xx

xxi

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

97

98

99

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

193

194

195

196

197

199

201

203

205

207

208

209

210

211

212

213

215

216

217

219

220

221

222

223

224

225

227

228

229

230

231

232

233

234

235

Technological Prospects and Social Applications Set

coordinated by

Bruno Salgues

Volume 4

Societal Responsibility of Artificial Intelligence

Towards an Ethical and Eco-responsible AI

Jérôme Béranger

First published 2021 in Great Britain and the United States 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 Ltd

27-37 St George’s Road

London SW19 4EU

UK

www.iste.co.uk

John Wiley & Sons, Inc.

111 River Street

Hoboken, NJ 07030

USA

www.wiley.com

© ISTE Ltd 2021

The rights of Jérôme Béranger 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 Control Number: 2021931051

British Library Cataloguing-in-Publication Data

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

ISBN 978-1-78630-694-4

Acknowledgments

This book is first and foremost the story of a professional human adventure that has been the fruit of much support and exciting interactions over the last few years.

In particular, I wish to thank the UMR 1295 BIOETHICS team of the Inserm/Université Paul Sabatier de Toulouse mixed research unit, and the UNESCO chair on “Ethics, Science, and Society” of the Université Fédérale de Toulouse for their trust and support in my research work on the ethical approach to AI.

Without all the people I met, this book would never have seen the light of day! I thank them very much.

Introduction

To the great promises that AI brings, it must meet a great human responsibility.

The digital world is characterized by its immediacy, its density of information, its omnipresence, in contrast to the concrete world of things. Now, with the multiplication of means of connection, the decrease in technology costs, the new capacities of data collection and algorithmic processing, we realize that we can communicate elements of our environment that were silent until now. We are witnessing the multifaceted development of new information and communication technologies (NICTs), illustrated by the emergence of technologies associated with Big Data; connected objects; algorithms; nanotechnology, biotechnology, information technology, and cognitive science (NBIC); blockchain; artificial intelligence (AI); virtual and augmented reality; and even quantum computing. AI is developing at an extremely rapid pace. We should expect to see significant changes in our society as AI systems become embedded in many aspects of our lives.

This multifaceted digital phenomenon is bringing different universes together by adding the speed, intelligence, and ubiquity of digital technology to the objects associated with these NICTs. Major developments related to AI in healthcare, autonomous vehicles, cybersecurity, education, home and service robots are improving the quality and comfort of our lives every day. Now, AI is fundamental to address many of the major challenges facing humanity, such as climate change, global health and well-being, natural resource development, and reliable and sustainable legal and democratic systems. This technology is, therefore, changing the way we live, consume, function and work. This is illustrated by a disruption with the past in the relationship and link that each person has with his or her neighbors. From then on, these interactions force the system to rethink each human activity. This is the beginning of a silent but very present revolution that is happening right before our eyes. A new era of change and disruption where survival inevitably requires reactivity, adaptability, creativity and, therefore, innovation.

Consequently, this technoscientific context is conducive to the development of an increasingly important international cultural and intellectual movement, namely transhumanism, whose objective is to improve the physical and mental characteristics of human beings by relying on biotechnologies and other emerging technologies. This current of thought considers that certain states of the human condition such as illness, disability, pain, aging and death are not fatal in themselves and can be corrected or even eliminated.

Thus, technological revolutions have enabled a change of scale in the exploitation of digital data, particularly in the field of genetics. They can be produced in large quantities, in an increasingly precise manner and preserved over an indefinite period of time. It can be observed that advances in computer science have made it possible, through the creation of specific programs, for databases to be interoperable, thus allowing for the fusion of data from various and multiple sources. To this, we can add the development of new ways of accessing data, in particular through the multiplication of data sources of all kinds. Crowdsourcing1 is becoming one of the new devices allowing easy access, in real time, to digital data in order to develop research (Khare et al. 2015).

ALGORITHMIC PROCESSING.–

Algorithmic processing is a finite and unambiguous sequence of operations or instructions to solve a problem or obtain a result. Algorithms are found today in many applications such as computer operation, cryptography, information routing, resource planning, and optimal use of resources, image processing, word processing and so on. An algorithm is a general method for solving a set of problems. It is said to be correct when, for each instance of the problem, it ends up producing the correct output, i.e. it solves the problem.

BIG DATA.–

Big Data, or megadata, sometimes referred to as massive data, refers to data sets that become so large that they are difficult to make use of with traditional database or information management tools. The term Big Data refers to a new discipline at the crossroads of several sectors such as technology, statistics, databases and business (marketing, finance, health, human resources, etc.). This phenomenon can be defined according to seven characteristics, the 7Vs (volume, variety, velocity, veracity, visualization, variability, value).

BLOCKCHAIN.–

Computer “block chain” is protected against any modification, each of which contains the identifier of its predecessor. The blockchain records a set of data such as a date, a cryptographic signature associated with the sender and a whole set of other specific elements. All these exchanges can be traced, consulted and downloaded free of charge on the Internet, by anyone who wishes to check the validity and non-falsification of the database in real time. The major advantage of this device is the ability to store a proof of information with each transaction in order to be able to prove later and at any moment the existence and content of this original information at a given moment. Its mission is, therefore, to create trust by protocolizing a digital asset or database by making it auditable.

CROWDSOURCING.–

A practice that corresponds to appealing to the general public or consumers to propose and create elements of the marketing policy (brand choice, slogan creation, video creation, product ideation/co-creation, etc.) or even to carry out marketing services. Within the framework of crowdsourcing, professional or amateur service providers can then be rewarded, remunerated or sometimes only valued when their creations are chosen by the advertiser or sometimes simply for their participation effort. Crowdsourcing has especially developed with the Internet, which favors the soliciting consumers or freelancers through specialized platforms.

AI appears as an essential evolution in the processing of digital information. It represents the part of computing dedicated to the automation of intelligent behaviors. This approach is the search for ways to endow computer systems with intellectual capacities comparable to those of human beings. AI must be capable of learning, adapting and changing its behavior.

The idea of elaborating autonomous machines probably dates back to Greek antiquity with the automatons built by Hephaestus, reported notably in the Iliad (Marcinkowski and Wilgaux 2004). For Brian Krzanich, President and CEO of Intel (the world’s leading microprocessor manufacturer), AI is not only the next tidal wave in computing, but also the next major turning point in the history of humankind. It does not represent a classic computer program: it is more educated than programmed. It is clear that the AI lawsuit has mixed fantasy, science fiction and long-term futurology, forgetting even the basic definitions of the latter.

Thus, the concept of AI2 is to develop computer programs capable of performing tasks performed by humans requiring learning, memory organization, and reasoning. The objective is to give notions of rationality, reasoning and perception (e.g. visual) functions to control a robot in an unknown environment. Its popularity is associated with new techniques, such as deep learning, which gives a program the possibility to learn how to represent the world because of a network of virtual neurons that perform each of the elementary calculations, in a similar way to our brain.

DEEP LEARNING.–

This algorithmic system has been used for more than 20 years for different actions in the form of neural networks, in particular to “learn”. A neuron represents a simple function that takes different inputs and calculates its result, which it sends to different outputs. These neurons are mainly structured and organized in layers. The first layer uses almost raw data and the last layer will generate a result. The more layers there are, the greater the learning and performance capacity will be. One can take the example of character recognition from handwriting. The first layer will take into account all the pixels that make up a written character – for example, a letter or a number – and each neuron will have a few pixels to analyze. The last layer will indicate “it’s a T with a probability of 0.8” or “it’s an I with a probability of 0.3”. A backpropagation operation is performed from the final result to remodify the parameters of each neuron.

The machine is programmed to “learn to learn”. AI does not exist to replace people, but to complement, assist, optimize and extend human capabilities. There are two types of AI:

– weak AI: its objective is to rid people of tedious tasks, using a computer program reproducing a specific behavior. This AI is fast to program, very powerful, but without any possibility of evolution. It is the current AI;

– strong AI: its objective is to build increasingly autonomous systems, or algorithms capable of solving problems. It is the most similar approach to human behavior. This AI learns or adapts very easily. Thanks to algorithmic feedback loops, the machine can modify its internal parameters used to manage the representation of each stratum from the representation of the previous stratum. These strata of functionalities are learned by the machine itself and not by humans. From this postulate, we can say that the machine becomes autonomous and intelligent, by constructing its own “computerization” structures and relying on axiomatic decisions. It is the future AI that should be developed in about 10 years.

WEAK AI.–

Weak AI or narrow AI simulates specific cognitive abilities such as natural language comprehension, speech recognition or driving. It only performs tasks for which it is programmed. It is therefore highly specialized. It is a machine for which the physical world is somewhat enigmatic, even ghostly, if it perceives it at all. It does not even have any awareness of time. This AI is unintelligent and works only on the basis of scenarios pre-established by designers and developers.

STRONG AI.–

Artificial general intelligence (AGI) or strong AI has similar – and even superior – reasoning abilities to those of human beings. It is endowed with capabilities not limited to certain areas or tasks. It reproduces or aims to reproduce a mind, or even a consciousness, on a machine. That is to say, an evolutionary machine with its own reasoning and consciousness, capable in particular of independently elaborating strategies and/or decisions that go beyond human beings in order to understand them so as to help them (in the best of cases) or to deceive or even destroy them (in the worst of cases).

From a general point of view, AI can be illustrated as an algorithmic matrix that aims to “justly or coldly” optimize decisions. Naturally, the morality or fairness of this judgment is not predefined, but depends, on the one hand, on the way in which the rules are learned (the objective criterion that has been chosen), and, on the other hand, on the way in which the learning sample has been constructed. The choice of the mathematical rules used to create the model is crucial. Just like the human functioning that analyzes a situation before changing one’s behavior, AI allows the machine to learn from its own results to modify its programming. This technology already exists in many applications like on our smartphones, and should soon be extended to all areas of daily life: from medicine to the autonomous car, through artistic creation, mass distribution, or the fight against crime and terrorism. Machine learning not only offers the opportunity to automatically make use of large amounts of data and identify habits in consumer behavior. Now, we can also actuate these data.

MACHINE LEARNING.–

Machine learning concerns the design, analysis, development and implementation of methods that allow a machine (in the broadest sense) to evolve through a systematic process, and, thus, perform tasks that are difficult or impossible to perform by more traditional algorithmic means. The algorithms used allow, to a certain extent, a computer-controlled (possibly a robot) or computer-assisted system to adapt its analyses and response behaviors based on the analysis of empirical data from a database or sensors.

In our view, adopting the machine learning method is no longer just a utility, but rather a necessity. Thus, in light of the digital transition and this “war of intelligences” (Alexandre 2017), companies will be the target of a major transformation and will invest in AI applications in order to:

– increase human expertise via virtual assistance programs;

– optimize certain products and services;

– bring new perspectives in R&D through the evolution of self-learning systems.

Therefore, AI holds great promise, but also strong fears, hazards and dangers that must be corrected or even removed, to ensure an implementation that is in accordance with the legal framework, moral values and ethical principles, and the common good. The conflicts in question can be very varied. Indeed, machines like robotic assistants ultimately ignore the concepts of good and evil. They need to be taught everything. Autonomous cars are likely to involve us in accidents or dangerous situations. Some conversational agents may insult or give bad advice to individuals and not be kind to them.

Thus, even if today, ethical recommendations have little impact on the functional scope of AI and introduce an additional level of complexity in the design of self-learning systems, it becomes essential, in the future, to design and integrate ethical criteria around digital projects related to AI.

Several standards dealing with algorithmic systems, transparency, privacy, confidentiality, impartiality and more generally with the development of ethical systems have been developed by professional associations such as the IEEE (Institute of Electrical and Electronics Engineers) and the IETF (Internet Engineering Task Force)3.

To this can be added documents focusing on ethical principles related to AI, such as:

– the Asilomar AI Principles, developed at the Future of Life Institute, in collaboration with attendees of the high-level Asilomar conference of January 2017 (hereafter “Asilomar” refers to Asilomar AI Principles, 2017);

– the ethical principles proposed in the Declaration on Artificial Intelligence, Robotics and Autonomous Systems, published by the European Group on Ethics in Science and New Technologies of the European Commission, in March 2018;

– the principles set out by the High-Level Expert Group on AI, via a report entitled “Ethics Guidelines for Trustworthy AI”, for the European Commission, December 18, 2018;

– the Montreal Declaration for AI, developed at the University of Montreal, following the Forum on the Socially Responsible Development of AI of November 2017 (hereafter “Montreal” refers to Montreal Declaration, 2017);

– best practices in AI of the

Partnership on AI

, the multi-stakeholder organization – composed of academics, researchers, civil society organizations, companies building and utilizing AI academics, researchers, civil society organizations and companies building and utilizing AI – that, in 2018, studied and formulated best practices in AI technologies. The objective was to improve public understanding of AI and to serve as an open platform for discussion and engagement on AI and its influences on individuals and society;

– the “five fundamental principles for an AI code”, proposed in paragraph 417 of the UK House of Lords Artificial Intelligence Committee’s report, “AI in the UK: Ready, Willing and Able”, published in April 2018 (hereafter “AIUK” refers to House of Lords, 2018);

– the ethical charter drawn up by the European Commission for the Efficiency of Justice (CEPEJ) on the use of AI in judicial systems and their environment. It is the first European text setting out ethical principles relating to the use of AI in judicial systems (see Appendix 1);

– the ethical principles of Luciano Floridi

et al

. in their article entitled “AI4People – An Ethical Framework for a Good AI Society: Opportunities, Risks, Principles, and Recommendations”,

Minds and Machines

, December 2018;

– the OPECST (

Office parlementaire d’évaluation des choix scientifiques et technologiques

) report (De Ganay and Gillot 2017);

– the six practical recommendations of the report of the CNIL

(Commission nationale de l’information et des libertés

)

4

on the ethical issues of algorithms and AI, drafted in 2017 (see Appendix 2);

– the report published by the French member of parliament Cédric Villani (2018) on AI;

– the Declaration on Ethics and Data Protection in the Artificial Intelligence Sector, at the 40th International Conference of Data Protection and Privacy Commissioners (ICDPPC), Tuesday, October 23, 2018, in Brussels;

– the seven guidelines

5

developed by the European High Level Expert Group on AI, published on April 8, 2019 by the European Commission;

– the five principles set out in the OECD Council Recommendation on the development, implementation and use of AI, adopted on May 22, 2019, by the Council to OECD Ministers

6

.

What is the best practice of ethical frameworks, regulations, technical standards, and best practices that are environmentally sustainable and socially acceptable? It is clear that these shared frameworks do not guarantee success. Mistakes and illegal behavior continue to occur. But their availability requires a clear and precise idea of what needs to be done and how to evaluate competing solutions.

This diversity of approaches and initiatives on the subject reflects the major challenge of establishing a common framework for ethical governance of AI. This raises a delicate and decisive question: how should the ethical governance of AI be defined, or by which characteristics? What are the “measurable” values, translating notions of loyalty, responsibility, trust and thus ethics applied to algorithmic decisions when they are the consequence or the result of a prediction?

It is from this vision of universalization that we felt the need to write this book around the framework of AI applicable to all. As a result, we have developed a moral framework to support digital AI projects by observing a number of requirements, recommendations and rules, elaborated, verified and discussed at each stage of design, implementation and use. This allowed us to design ethical criteria, according to our determinants, both essential and universal, based on the principle of Ethics by Design7 or Human Rights by Design to move toward a totally innovative principle of Ethics by Evolution that we will develop throughout this book. The objective is to achieve AI that is safer, more secure and better adapted to our needs, both ethical and human, over time. This will help optimize our ability to monitor progress against criteria of sustainability and social cohesion. AI is, therefore, not an end in itself, but rather a means to increase individual and societal well-being.

ETHICS BY DESIGN.–

An approach that integrates ethical requirements and recommendations from the design of NICTs.

ETHICS BY EVOLUTION.–

It is an approach that incorporates recommendations and ethical rules, in an evolutionary manner over time, throughout the lifecycle of NICTs, i.e. until its implementation and evolutionary use.

This book is intended to categorize ethical issues related to the digital environment, both from the point of view of the user and the designer of digital solutions and/or services. It invites reflection (what questions businesses can ask themselves about digital ethics) and suggests avenues for action. It is an approach that aims to provide guidelines to bring out the values that we want to collectively put forward to help legislators to formulate laws that will build a framework for AI. This repository is not exhaustive. It is intended to be general, open to all contributions and evolving. It must be regularly updated to ensure its consistency and constant relevance as the digital environment and our technological knowledge evolves. It is intended as a reminder of the company’s regulatory duties, which precisely define what is permitted or prohibited, and the sanctions that apply. The company has an obligation to comply, and this does not concern the area of ethics. However, the means by which it complies can be the subject of ethical reflection.

Finally, this book is addressed to all stakeholders involved in the development, deployment or use of AI, including organizations, companies, public services, researchers, individuals or other entities. This document should, therefore, be considered as the first building block of a discussion between these different actors toward an ethical, responsible, trustworthy AI aimed at protecting and serving in a beneficial way individuals and the common good for a better adoption at the global level.

1

In France, crowdsourcing is defined according to the

Commission générale de terminologie et de néologie

(2014) as the “mode of completion of a project or a product calling for contributions from a large number of people, generally Internet users”.

JORF

, 0179(91), 12995.

2

ISO 2382-28:1995 defines artificial intelligence as “the capability of a functional unit to perform functions that are generally associated with human intelligence, such as reasoning and learning”.

3

IEEE P7000:

Model Process for Addressing Ethical Concerns During System Design

; IEEE P7001:

Transparency of Autonomous Systems

; IEEE P7002:

Data Privacy Process

; IEEE P7003:

Algorithmic Bias Considerations

; IETF

Research into Human Rights Protocol Considerations

draft.

4

CNIL (2017).

Comment permettre à l’homme de garder la main ? Les enjeux éthiques des algorithmes et de l’intelligence artificielle

. Summary report of the public debate led by the CNIL in the context of the mission of ethical reflection entrusted by law for a digital Republic.

5

These seven essential requirements include human factor and human control, technical robustness and security, privacy and data governance, transparency, diversity, non-discrimination and equity, societal and environmental well-being, and accountability.

6

On May 22, 2019, through the OECD Council of Ministers, 42 countries (the 36 OECD countries and Argentina, Brazil, Colombia, Costa Rica, Peru, and Romania) adopted the principles set out in the OECD Recommendation on AI, making it the first intergovernmental agreement to stimulate innovation and build confidence in AI by promoting a responsible approach to trusted AI, while ensuring respect for human rights and democratic values.

7

This consists of integrating ethical rules and requirements from the design and learning of these NICTs, prohibiting direct or indirect damage to the fundamental values protected by the conventions.

1Societal and Moral Questioning Around AI and Its Ecosystem

From autonomous cars to facial or voice recognition, artificial intelligence (AI) has developed and structured itself in a spectacular way over the last 5 years and is now part of our daily life and close environment. The widespread use of algorithmic applications feeds our imaginations, hypnotized by the promise of a better world, where the computing power of machines could reduce or even eliminate illnesses, accidents and crimes. At the same time, a growing doubt about AI is beginning to develop, portraying the technology and its exponential progression as a potential danger to the survival of humanity. Indeed, by the end of 2017, entrepreneur Elon Musk – the charismatic head of Tesla and Space X, among others – believed that efforts to make AI safe had a 5–10% chance of success. In doing so, he reaffirmed his earlier prediction that there was a risk that something very dangerous was going to happen within the next 5–10 years. At the heart of the concerns is technological singularity, a concept that predicts a runaway technological progress that could lead to the advent of a superhuman AI that would have autonomous capabilities to improve and evolve.

Moreover, there is no development of AI without the exploitation of gigantic volumes of data (Big Data). Indeed, we have to keep in mind that the computer without information can neither learn nor automate human action. AI stores the information that humans decide to give it. Failing to limit the collection of data, it would be appropriate to control its use and protection. As a result, the acquisition, storage, consumption and management of this Big Data are two decisive requirements for any contemporary society. So, with the digital revocation of AI and Big Data, all businesses and organizations have become aware of the potential that lies before them.

Now they want to highlight this relevant information and take full advantage of it. But how do you leverage accessible information while ensuring that you have a high-performance ecosystem in place to store, analyze and develop it? The emergence of algorithmic systems also raises the anxiety of a world guided and controlled by digital logic. To what extent can we leave the control of our contemporary societies to algorithms and those who design them? How can we guarantee the confidentiality of our private lives from the growing appetite of machines fed by the collection of personal data? How do we prepare for the upheavals and consequences that AI will bring about in many professional sectors? These are all questions that will be the subject of reflection in Chapter 2.

In order to pragmatically apprehend and understand these issues centered on human dignity and its fulfillment, we introduce four challenges and perspectives offered to society by AI (Floridi et al. 2018):

– whom can we become (autonomous self-realization): AI can assist in self-realization, i.e. the ability of individuals to develop in terms of their own characteristics, interests, potential abilities or skills, aspirations and projects;

– what we can do: AI allows us to improve and multiply the possibilities of human representation. Responsibility is paramount, given the kind of AI we develop, how we use it, and how much we share with all of its advantages and benefits. AI applications could help, if designed effectively, amplify and strengthen distributed and shared moral systems;

– what we can achieve (individual and societal abilities): if we rely on the use of AI-related technologies to increase our work capacities, we will be able to delegate certain tasks and especially decisions concerning autonomous systems that must remain at least in part subject to supervision and human choice. It is, therefore, becoming essential to find a balance between, on the one hand, the pursuit of the ambitious prospects and opportunities offered by AI to improve human life and what we can achieve, and, on the other hand, to ensure that we remain masters of these major developments and their impact on human society;

– how we can interact with each other and with the world (society, cohesion): AI can go a long way in dealing with such complex coordination, supporting more societal cohesion and collaboration, without undermining human dignity and without eroding human self-determination.

1.1. Use cases of AI

Now, AI applications are invading all sectors of activity and professional spheres of the company. Algorithmic uses are tirelessly multiplying and diversifying a little more every day. In each case, AI can be used to enhance human nature and its performance, creating actual opportunities that must be seized and well used (see Table 1.1).

Table 1.1.AI use cases by industry

Industry

AI use cases

Cities and local authorities

Increase user access to public services:

– free up agent time by freeing them from repetitive tasks;

– guarantee universal access to public service by breaking down the language barrier.

Simplify citizens’ lives and experiences: ensuring the efficiency of shared services.

Optimize the management of the public budget: make the funding coincide with the actual consumption of goods.

Education

Better meet the needs of students:

– prevent school and academic dropout;

– support students outside of the institution;

– support students in their choices.

Transform learning:

– promote learning to read;

– propose personalized courses.

Assist teachers:

– allow teachers to spend as much time as possible with their students;

– offer teachers feedback on their courses.

Change the report to information.

Streamline registration procedures.

Banking and insurance

In customer relationships:

– develop commercial relationship;

– save time;

– speak the same language as the client.

Reduce risk and fraud: fraud identification and anti-money laundering.

Create new business models:

– combine different businesses through data aggregation;

– manage large numbers of investment simulations;

– the creation of new products to support healthcare professionals in their daily practices.

Health

AI for better public health: detection and treatment:

– revolutionize medical imaging;

– harmonize care processes and support the doctor in their application;

– generate alerts and reminders to healthcare professionals and/or patients;

– review therapy and care planning;

– recognize medical images and interpret them (radiology, ophthalmology, dermatology, etc.);

– assist in paramedical care (paramedical humanoid robot);

– assist in medical decision-making and establish predictive analyses via a diagnostic assistant;

– enable communication interfaces between patients and healthcare professionals via a conversational agent (chat-bots) (conversational oncology);

– monitor patients in real time and adjust their treatments to their individual situation;

– make an earlier and more accurate diagnosis;

– access to new knowledge;

– improve the flow of the city hospital route;

– reduce costs and pool resources.

Retail

Lead the customer to the store:

– multiply access points: voice;

– multiply access points: images.

Transform the buying journey and improve customer relations:

– make the customer’s journey through the store more fluid;

– provide an interactive customer experience;

– adapt products to customers’ needs and desires;

– use facial recognition for various actions toward the customer;

– offer customers the assistance of a personal digital assistant (PDA).

Empower employees to do more by being more efficient:

– facilitate the maintenance of in-store shelves;

– alert employees when a customer needs them;

– proposals targeted to the user.

Optimize operations:

– optimize the organization of stores and promote sales;

– use customer data;

– optimize inventory and inventory management;

– optimize the delivery of orders and reduce costs.

Manufacturing industry

Optimize the production chain:

– optimize the flow of production lines;

– optimize the quality of the production lines.

Improve the maintenance process:

– predictive maintenance;

– facilitate the work of maintenance agents.

Strengthen employee safety.

Make relations with suppliers more fluid.

Obtain a better knowledge of the customer in order to better serve him or her.

Corporate Finance – CFO

Make sophisticated and reliable predictions:

– obtain forecasts of the company’s financial data and guide strategic decision-making;

– determine employee bonuses.

Free up human time (day/human):

– answer questions from business teams;

– manage the billing process;

– check expense reports.

Marketing

AI for creative campaigns.

Evaluate the performance of campaigns with a new level of accuracy.

Promise a personalized experience.

Constantly improve based on real-time feedback.

A fluid client interface thanks to cognitive services.

Sales manager and customer relations

AI at the heart of the customer journey:

– detect buying signals;

– better manage the pace of customer interactions;

– in store, detect abnormal behavior or optimize sales staff interventions;

– automate information retrieval and call center operations.

Gain time and efficiency in day-to-day work.

Human resources (HR)

Improve the recruitment process:

– attract candidates through language analysis;

– facilitate the application process;

– identify relevant candidates.

Giving employees the means to develop:

– make interactions more fluid;

– offer adapted and personalized training;

– enable employees to find their next job with the company.

Relieve HR managers of certain tasks:

– contribute to risk and compliance analysis;

– predict recruitment needs.

CHATBOTS.–

A software robot that can talk with an individual or consumer through an automated conversation service that can be carried out through decision trees or by an ability to process natural language.

Finally, we can envision three logics or ways in which AI will impact employment and the field of activity:

– a substitutive logic: it mainly concerns jobs that are not very highly qualified and implies an accompaniment toward new jobs for the collaborators and workers concerned;

– a rationalizing logic: this involves low-value-added jobs which, thanks to AI, could become less burdensome, for example, workers who work in assembly lines in factories;

– a capability logic: AI comes to enhance human capabilities by refocusing the activity on its added value and strengthening it (better performing and more relevant employees, distinctive skills that can be enhanced, etc.). Administrative and analytical tasks are then transferred to the autonomous algorithmic system to focus on human added value.

1.2. Digital environment

For the past 20 years or so, we have been living in a world where data are constantly multiplying, where visions and analyses have become infinite and where everything has become a sum of singularities and values whose structures seek to understand how to extract them. These large volumes of unpublished data are helping to store and build new knowledge, new perceptions and, therefore, new opportunities. We are now in an era of convergence between data, which can all become homogeneous, digitized and integrable, and with more correlation of senses. This is the digitization roll-out of the world, where databases and tools for storing and exploiting data on a large scale have been completely rethought and improved, considerably enhancing their operational performance. For more than a decade, online exchange platforms on the Internet have become the places where information, communication, knowledge and sociability converge. Digital culture represents the continuous path between the concrete and the conceptual, between the real and the virtual, adjusted by the evolutions of the digital environment. We are in the middle of an authoritarianism of immediacy, instantaneity and acceleration of the rhythm of life. This is illustrated by new Data Driven-centered approaches, where we visualize more than we model and where quantity takes precedence over quality. The consequence of this “datafication” is to provide the conditions and the means for governments and businesses to map society in a quantifiable and analyzable way, for an in-depth analysis of the reality that permeates people’s daily lives, and even their thoughts.

DATA DRIVEN.–

An approach that involves making strategic decisions based on data analysis and interpretation. This approach allows for data to be examined and organized in order to better understand its consumers and customers. “Data driven” will, therefore, allow an organization to contextualize and/or personalize the message to its prospects and customers.

By its characteristics, AI irremediably leads to a considerable conceptual change around its digital ecosystem. Data warehouses are no longer at the center of the world. Many repositories and specialized tools support applications or new forms of analysis. Increasingly, data are coming from sources outside the infrastructure through application programming interface (APIs). As a result, the company that processes these data is more like a distributed supply chain.

API.–

In computer science, an API is a standardized set of classes, methods or functions that serves as a front end through which one piece of software provides services to other pieces of software. Software such as operating systems, database management systems, programming languages or application servers have an API.

Because of its systemic dimension, the digital revolution is causing major upheavals within companies and society. With the same disruptive force as the industrial revolution, it is transforming the business model, organization, corporate culture and strategy, and management style. As a result, companies using Big Data via AI are, therefore, faced with a new ecosystem that can be divided into nine segments (Kepeklian and Wibaux 2012, see Table 1.2).

Table 1.2.Digital ecosystem of Big Data operated by AI

Network of partnersHostDatacenterHPCManufacturerCloud operator

Key activitiesAdviceCalculationStorage

OfferCollectingProcessing logsVertical applicationsAnalyzing VisualizingInterpretingStructuringStoring

Client relationsSelf-service Training Support

Client segmentAdministrationMedia IndustryBankingHealthDistribution Etc.

Key resourcesSizePerformanceSpace

Distribution channelsB2BA2B

Cost structurePlatform maintenanceSubscriptionDevelopmentExploitation

IncomeflowData valorizationSales/re-salesBatch/transactlonsIP, licenses, rentals

CLOUD COMPUTING.–

Cloud computing is an infrastructure in which computing power and storage are managed by remote servers to which users connect via a secure Internet connection in order to deliver faster innovation, flexible resources and economies of scale. The desktop or laptop computer, cell phone, touch-tablet and other connected devices become access points for running applications or consulting data that are hosted on servers. The Cloud is also characterized by its flexibility, which enables vendors to automatically adapt storage capacity and computing power to user needs.

In addition, this digital environment is subject to many changes, both internal, particularly with:

– the explosion of dematerialized services and large volumes of data (mostly unstructured) coming in particular from the activity of Internet users and infrastructures;

– the decentralized ecosystem of companies and the heterogeneity of internal security levels justifying the implementation of data protection standards.

And of an external nature, in particular:

– the convergence of professional and domestic uses of digital applications. Indeed, social networks, discussion forums, Wi-Fi, blogs, instant messaging, Wikipedia, etc., are uses that are often incompatible with applications and needs within a company. This is why more and more professional infrastructures are establishing codes of good practice, rules of use and even ethical charters describing the value of these new information and communication technologies (NICTs) for the company;