The Algorithmic Code of Ethics: Ethics at the Bedside of the Digital Revolution

By Jérôme Béranger

The technical progress illustrated by the development of Artificial Intelligence (AI), Big Data technologies, the Internet of Things (IoT), online platforms, NBICs, autonomous expert systems, and the Blockchain let appear the possibility of a new world and the emergence of a fourth industrial revolution centered around digital data. Therefore, the advent of digital and its omnipresence in our modern society create a growing need to lay ethical benchmarks against this new religion of data, the "dataisme".

• Language: English
• Publisher: Wiley
• Release date: Oct 8, 2018
• ISBN: 9781119549673

Jérôme Béranger

Jérôme Béranger is a senior consultant and associate researcher for Keosys. His research focuses on ethics and the sociology of information systems intended for e-health, m-health and Big Data.

Book preview

Foreword

Welcome to a new world where mankind joins forces with machines through digital technology. To understand this new world, there is only one guide: the book of Jérôme Béranger.

Both lighthearted and serious, clear and complex, a source of hope as well as concern, the author’s reflections allow us to navigate between the questions and interests that digital technology and its applications raise.

This reconciliation of contradictions is easily achieved as the author helps the reader to ask the right questions by suggesting that they refer to a Code of Ethics.

Digital technology allegedly originated in nature where, in fine, all becomes one. Following the example of a matrix, nature is allegedly only a visual envelope which reflects a virtual reality. How can we then make sense of it, if not by relying on what is most fundamental and common to us? The answer is: ethics. As such, the author suggests inserting ethics into the meandering of digital technology, in order to come up with an algorithmic ethics, an applied ethics and a renewed ethics; an ambition which is sometimes difficult to consider for a lawyer with a Cartesian mind. To come to terms with it, returning to the word’s definition is necessary. Thus, French dictionaries commonly define ethics as all the principles of morality that are the basis for a person’s behavior. Ethics is thus generated by a moral that people question, challenge, contradict, and then accept and apply in order to comply with it. It is therefore a reference specific to each person.

Consequently, can codifying this ethics and creating an algorithmic ethics be considered?

With a strong, scientific demonstration, using mathematical and philosophical concepts, the author convinces the reader as to the need for a regulation of the digital world. This regulation cannot be exclusively legal; it must also be ethical. Thus, Jérôme Béranger suggests a digital governance laying the foundations for a responsible and empowered society whose foundation would be an algorithmic code of ethics.

Along the lines of a social pact within a digital society, Jérôme Béranger thus humbly lays down the foundations for our future societal model.

Lina WILLIATTE

Professor of Law – Université Catholique de Lille

Lawyer at the Lille bar – Cabinet WT Avocat

Acknowledgments

I wish to seize this literary opportunity to express my deep appreciation and my sincere gratitude to:

– The ADEL corporation (Frédéric, Conrad, Guillaume and Jérôme), JMC Investment (Jean-Marie and Tom) and Team 4 of the UMR 1027 (public health and epidemiology research center) of the INSERM (French National Institute of Health and Medical Research) mixed unit/Paul Sabatier University (in particular Emmanuelle) for their support and trust in my research, as well as for providing me with the best possible conditions – in terms of material and human resources – in order to develop my innovating and forward-thinking vision of a digital ethics;

– Cédric Villani and Gilles Babinet, who have supported me from the very beginning of my work and reflections on the subject;

– Lina Williatte, who was so kind as to write the foreword of this book;

– All persons who have, directly or indirectly, taken part in the completion of this work;

– My family (Madeleine, Jean-Noël, Stéphanie, Christophe, etc.) and my friends (Yannick, Marie-Ève, Élodie, Marjorie, Mathieu, Danaé, Emma, Nicolas, Marie, etc.) for their encouragements and patience. I dedicate this book to them…

Without all these people, this book would never have seen the light of day!

Introduction

Everything changes, nothing remains without change.

Buddha

In nature, everything always has a reason. If you understand this reason, you do not need any more experience.

Leonardo da Vinci

It has become appallingly obvious that our technology has exceeded our humanity.

Albert Einstein

Information is the resolution of uncertainty.

Claude Shannon

I.1. Preamble

In the 1950s, IBM sought to introduce the word computer in France. Philosopher Jacques Perret, mandated by the American giant, came up with the idea of using an ancient Latin word from the Middle Ages: ordinateur. This term referred to a quality that the Church Fathers assigned to God – Deus Ordinator – meaning God Orderer [ALI 17].

Over half a century later and after the emergence of algorithmic systems, one could wonder if the use of this name was premonitory. Indeed, biblical texts say that God is everywhere, in each one of us […] and guides our choices. Two thousand years later, this quote is still relevant, with an aging God who has taken a digital form: Algorithm is everywhere, in each one of us […] and guides our choices!

For the historian Yuval Noah Harari [HAR 17], the 21st Century will witness the emergence of a new religion focused on data, dataism, in order to fill a spiritual void, and whose premise claims that the universe consists of a data stream and that the value of each phenomenon or entity is determined by its contribution to data processing.

By inventing the binary code a few years ago, human beings had certainly not considered that they would become themselves a series of 0s and 1s. In fact, nowadays, algorithms are omnipresent within us and our environment in a natural and artificial manner, whilst governing us directly or indirectly. Information is everywhere and the algorithm enables mankind to interact with this information environment. In 2014, GSMA Intelligence announced that there were more connected devices than human beings on the planet. All the ins and outs of theirs uses are not known.

The datafication, algorithmization, or even gamification¹ of our society inevitably result in a dematerialization of relationships, a disintermediation and a transfer from property value to use value. This data society is no longer built on meaningful knowledge, but on data aggregates, which contribute to the disruptive nature of our societies [STI 15].

Algorithms are now an integral part of the citizen’s daily life with the hierarchical organization of the information on billions of people’s consumption on Google (PageRank), the selection of information present on Facebook’s newsfeed (Edgerank), product recommendations (Amazon), the optimization and geolocation of movements, disease detection and so on. They represent the basic operation tool of most current websites and software, and the operating source of connected objects², information systems (IS), future digital ecosystems and technologies (big data, Artificial Intelligence (AI), platforming, IoT (Internet of Things), Blockchain, biometric system, virtual or even augmented reality, Nanotechnology, Biotechnology, Information Technology (IT) and Cognitive Science (NBIC), quantum computing, three-dimensional bioprinting³, etc.).

These algorithms are probably one of the most valued immaterial creations of our modern economy. Nearly all market participants base their model on their algorithms’ performance. All these algorithmic actions include within each of them a moral process containing their own value system, their own way of reading and understanding reality and its interrelationships, as well as their own subjectivity. Algorithms have no qualms about invading people’s privacy and studying their behavior. In these conditions, designing, implementing and using these algorithms cannot remain neutral due to their societal and economic impact. Consequently, this digital transformation introduces a change of contextual landmarks and perspectives focused on information and its financial valuation potential. From now on, the value of data is conditioned by a set of transversal parameters and criteria, such as its accessibility, integrity, reliability, legitimacy, purpose, confidentiality, and so on.

In our current digital world, compliance with the law and regulations is no longer sufficient. The ethical dimension of algorithmic data processing⁴ must be reviewed. Governing and regulating New Information and Communication Technologies (NICTs) must go way beyond the rigid, purely technological and normative aspect, in order to embrace the transversal, flexible and mobile dimension of algorithmic ethics. This is why it is essential to consider an in-depth, wide and multidisciplinary analysis, in order to provide recommendations and clarify the social debate as to how modern society should face this technological challenge.

Therefore, confronted with this change in paradigm of our society, how should mankind evolve? It is obviously complicated to project in a future which is, by definition, uncertain, especially since we are in a transition period marking the beginning of the Fourth Industrial Revolution in human history. The whole point, essence and purpose of this book are actually based on this issue. We are convinced that, on the basis of our ethical approach, we can provide some responses regarding the main course offered to us, along the path of our neo-Darwinian evolution.

This book is a follow-up to two previous books entitled Medical Information Systems Ethics [BÉR 15] and Big Data and Ethics [BÉR 16] published by ISTE. It is the last part of our trilogy on ethics in the digital ecosystem. Over these years of achievements and reflections, our line of reasoning evolved from digital ethics to algorithmic ethics and ended up with quantum ethics. Regardless of the name, form and application of our ethical vision on the subject, the intrinsic essence and content emerging are first and foremost what prevails. Our literary quest aims at pursuing our vigilance and our epistemic, civic and moral approach regarding the societal issues to which we are confronted, as they involve our future, and whose risks, challenges and limits, beyond an apparent semantic modernity, seem a priori obscure and meaningless.

We are fortunate to be at the beginning of a new era based on the digitalization of society. It is therefore essential to be able to seize this opportunity to define a human framework of confidence and outline a new moral space focused on ethical principles and the accountability of all concerned participants, by developing, implementing and using NICTs. This is what is called ethics by design!

Hence, the main goal of this book is to provide readers with the first building blocks in the interest of the people, which go with this digital revolution, in order to obtain a more meaningful and better harmony between the intentions, implementations and purposes of technological tools. The balance between the promise of useful innovation and the risk of algorithmic prejudice is at stake. Therefore, by highlighting a background focused on a new ethical and technical approach, our book aims, on the one hand, to provide the backbone for a substantial change in attitudes surrounding the digital environment of society and, on the other hand, to contribute to the intelligibility of this digital revolution and ponder the impact of our digital representations in terms of social, economic and industrial practices, construction of standards and relationship with civic action.

Finally, our reasoning about the establishment and integration of an algorithmic code of ethics will include two essential steps, which are both separate and complementary. The first stage will argue that Code is Ethics, following the example of the legal expert Lawrence Lessig’s famous quote that Code is Law [LES 99]. The author stresses that code has gradually been set up as an active regulator of behaviors and of the online dematerialized transactions of the collaborative economy which, de facto, supersedes the law. The encoder is generally solely responsible for their choices and is the only one who can determine whether this choice is ethical. For instance, this is the case for the quantity of data they will use to perform algorithmic processing, for which they will have to carefully think about the really useful data and encode the algorithm accordingly. Technology is likely to set the standard without any democratic debate. Its value also comes from the fact that code is a workspace for developers and, as such, reflects the culture of their team and company. The influence of computer code on our daily lives and the overriding of the law by the algorithm can be witnessed. Digital technology generates numerous ethical questions of its own, as it alone produces a regulatory system equivalent to that of the law.

Then, a second phase will demonstrate that Ethics is Code, adapting Primavera De Filippi and Samer Hassan’s expression [DEF 16]: Law is Code, suggesting that blockchain (distributed public register) could be the technical source making it possible for code to become a way of enforcing the law. This regulatory perspective of digital technology must be the subject of an ethics of attention, which could highlight dilemmas between the standards produced by digital technology and our own standards or value systems. However, this ethics is likely to be insufficient. Indeed if, as Lawrence Lessig [LES 99] has said, the code regulates, and if it regulates poorly, if the intentions of the programmers are not in line with a certain ethics, regulating or rebalancing this normative vision will be difficult a posteriori. Under these conditions, an autonomous platform or expert system could initiate technological solutions allowing the code to become a means of expression for ethics. As part of a neo-Darwinian approach, this platform, or even AI⁵, would be the embodiment of our idea of an algorithmic code of ethics, both scalable and universal. This is then called ethics by evolution!

Lastly, the particularity of this book is its focus on both NICTs animated by algorithms themselves and the best ethical requirements and recommendations regarding their design, development, use and control, while encouraging the latter.

I.2. Technical revolutions through time

Initially reflecting the circular movement of a star returning to its point of origin, the term revolution has changed with time and now refers to a radical break or change in our structural patterns and our framework for reflection. This semantic insight reflects the cyclical nature seen in a revolution through time, which results in different forms. If the industrial sector is associated with it, this revolution is defined by a collection of important changes which successively transformed industrial production devices and work, significantly affecting the employment, economy, consumption and environment. This definition clearly explains that the change in our vision of the world is the result of each Industrial Revolution and is currently taking place again.

Over the centuries, science has made the discovery of new energy resources and raw material possible, facilitating the appearance of more efficient machines and technological innovations, as well as novel production methods. Therefore, the technological evolution of mankind has been driven by new knowledge and innovations created, acquired and used by humans. This necessarily leads to a phenomenon of obsolescence⁶ of the actors’ knowledge, which is inevitably outdated and superseded by NICTs. Consequently, in a perpetually shifting and evolving world, it is essential that everyone avoids keeping the same bases and rather enriches them, in order to develop their expertise within the society. The different realms, the Western European countries and then the New World, have mainly sought, during each major phase of their history and through their multiple mutual influences (economic, political, spiritual, cultural, social, etc.), a great common and shared goal, whose structure, which is spatially and temporally quite homogeneous, provided the development of a society model each time [DOS 15].

For the record, it is worth remembering that Europe was the birth place of what we call the Western model up to the end of the 19th Century (in particular with its colonial development policy). Then, the United States appeared on the international scene (via its globalization policy) during the 20th Century. Finally, the latest projections and trends predict that the United States and various Asian countries, especially China, will extend their influence worldwide.

Technical progress led humankind to improve its industry throughout its history and increasingly rely on it, but also reinvent it as soon as new resources provided new technical means. Therefore, industry was the subject of qualitative and significant progress representative of its time, which can be associated with revolutions. Thus, we see that the cyclical history of mankind – which endlessly starts over – currently accepts four main industrial and technical revolutions, which are fully interdependent:

– The First Industrial Revolution took place at the end of the 18th Century in the United Kingdom and then France: it emerged with the advent of mechanization which resulted in industry becoming the cornerstone of the economy, sidelining agriculture. The invention of the steam engine associated with the massive extraction of coal contributed to a new essential energy (patented by James Watt in 1760) for the development of material, economic and human exchanges via the expansion of railways. The first railway steam engine was built and then patented in 1815 in the United Kingdom by George Stephenson. Other significant inventions, in particular in textiles (mechanized sewing machines, the Jacquard loom created by Joseph Jacquard in 1801) or in metallurgy (metal and cast iron work, use of coke and invention of puddlage) and in the steel industry (in France, the pont d’Austerlitz was the first metal bridge built in 1807), amplified this phenomenon of society industrialization.

We then witnessed a rapid development of economic growth driven by the unprecedented increase in production generated by major discoveries in the knowledge sector. Urban centers were gradually built and grew. The establishment of different companies in the same place made it possible to improve the profitability of productions. Technical innovations were always renewed, constantly increasing the industrial output. This therefore led to major economic and social disruptions. Finally, this First Industrial Revolution emerged in a specific context favorable to progress and during which the capitalist spirit increasingly developed. It is thus that, up to 1830, significant economic, technical and social changes reshaped Great Britain, before extending to Europe and the United States.

– The Second Industrial Revolution emerged at the end of the 19th Century in Germany and the United States: after an economic depression (Vienna crash) over several years (1873–1896), a Second Industrial Revolution was born. It started at the very end of the 19th Century and only ended in 1914 with the beginning of WWI.

This Industrial Revolution began again with the appearance of new energy sources (gas, electricity and oil) which made the creation of new technological innovations and inventions possible, such as the internal combustion engine, artificial textiles, the incandescent light bulb, the telephone (by Bell), the telegraph (by Morse), the automobile, the plane, and so on. These means of communication and transport facilitated the creation of international trade.

This was the beginning of a period where the economy and industry were based on new methods of production and management devised by Taylorism and Fordism, and from which the steel industry developed steel and aluminum. Likewise, the economy adopted a new model with the emergence of trusts, cartels and the increasing number of shareholders. Therefore, this period is marked by great innovations which facilitated the economic growth of countries and productivity gains.

Consequently, this Second Industrial Revolution contributed to the emergence of major technical progress, the creation of large companies, especially in the United States, the beginning of globalization and commercial (mainly international), financial and human (human migrations) development. It was accompanied by a reduction of inequalities in industrialized countries and a gradual increase of the workers’ standard of living.

– The Third Industrial Revolution emerged during the mid-20th Century in the United States: as before, it was the appearance of a new energy, nuclear generated electricity (Uranium), that marked the beginning of this new Industrial Revolution. This source of energy resulted in the development of revolutionary materials (silicones, ceramics, resins) and contributed worldwide to the structuring of new electronic means of transmission (micro-processor and transistor: see Information Technology 1.0⁷) and information technologies to automate production. This opened the way for component miniaturization, which then made electronics one of the main areas of the economy. This period was characterized by the production of ever smaller materials and tools, which had an impact and opened up new prospects, especially in the aerospace, aeronautic and biotechnology industries. Let us note that some experts consider that the history of electronics started with the invention of the vacuum tube by Fleming in 1904, followed by the invention of the triode by Lee de Forest in 1907.

It was in the United States, more specifically in California in Palo Alto, that this revolution really started at the beginning of the 1940s with the creation by William Hewlett and David Packard of the first company based on digital technology, which resulted in the emergence in 1971 of the Silicon Valley, the first global technopole. At the end of WWII, the engineer Vannevar Bush invented a memorizing machine storing microfilms. At mid-decade, in the United States, the Citizen-Band (or CB) activity appeared, which was the first involvement of amateurs in the field of telecommunications. In 1957, the USSR put the first artificial satellite into orbit, Sputnik 1. This stage was decisive in the age of telecommunications, especially for the implementation of the Internet a few decades later. The process of component miniaturization – as the image of the integrated circuit (or electronic chip) invented by Jack Kilby (of Texas Instrument), or the modem (by Bell) making it possible to disseminate binary data – relentlessly continued. This phenomenon increased production cost reduction, while programming languages were more and more elaborate via increasingly powerful algorithms. Computer commercialization had only just started and only concerned the business field.

Finally, in 1969, thanks to the research of Léonard Kleinrock (at the MIT) on the use of packet switching for data transfer, the ARPANET (Advanced Research Projects Agency Network) project was created. The implementation of this device within the US Department of Defense took place in the context of the Cold War with the USSR. The challenge of this system was to develop a structurally decentralized military telecommunications network able to operate despite the destruction of some devices or line breaks. Its development and use at the level of the whole society, which was never considered at the time, marked the arrival of the Internet and, along with it, the beginning of the Fourth Industrial Revolution or Digital Revolution, at the beginning of the 1990s.

– The Fourth Industrial Revolution took place at the end of the 20th Century in the United States and China: this revolution was directly associated with the emergence of the massive development of information technology, namely the fact that any information can be expressed through a combination of numbers (namely 0s and 1s). We consider that this Fourth Industrial Revolution appeared at the beginning of the 1990s, with the appearance of the most famous form of Internet today, the Web⁸.

Let us note that the philosopher Luciano Floridi [FLO 14] also referred to the Fourth Revolution in his book entitled The Fourth Revolution, in order to characterize this new ecosystem focused on digital information reconciling nature (physis) and technology (technè). The author assumed that human beings represent informative organisms (inforgs) [FLO 07] among others, which are not very different from material entities, natural or artificial agents, smart systems, or modified connected objects.

It can be noted that in 1991, China was one of the pioneers of this digital revolution with the design of the first TCP/IP⁹ network, called the TUNET, at Tsinghua University. Three years later, a first connection to the Internet was designed by linking the electron spectrometer of Beijing to the linear accelerator of Stanford University.

Today, we are right in the middle of the period characterized by a fusion of NICTs, which blurs the lines between physical, digital and biological spaces worldwide. Digitalization facilitates the construction of a new, virtual world from which it is possible to drive the physical sphere, especially thanks to the analysis of databases and big data. Therefore, data represents the raw material of this digital revolution. In fact, it is often compared to oil, which is itself the main element of the Second Industrial Revolution.

Thus, these industrial revolutions helped to develop new products and new technologies. This can be illustrated by the creation of personal computers (see Information Technology 1.0) (1980s), the exponential growth of Internet uses, such as websites, forums and social media (see Information Technology 2.0¹⁰) (1990s), and the appearance of connected objects, smartphones, tablets, the Industrial Internet of Things (IIoT), (see Information Technology 3.0¹¹) (2000s). For example, it can be said that NICTs, such as connected objects, play the same catalytic role for big data as chemistry and automobiles for oil.

This Fourth Revolution is the first one to directly impact and modify the everyday space–time landmarks of people. There is indeed a real digitalization and platformization of society (AI, Programmable Logic Controller (PLC), self-learning expert systems, biotechnologies, three dimensions (3D), and robots, Uber, Airbnb, BlaBlaCar, smart cities, etc.) (See Information Technology 4.0¹²) (2010s). In fact, all big cities in the world are starting to fully integrate the digital aspect. These smart cities are made possible through the installation and widespread usage of sensors (along with its inhabitants), as well as networking these sensors via telecommunications operators. The objective is to provide new services, improve the quality of life and security, and have a more sustainable city (urban planning, transport management, citizen integration, etc.). The installation of sensors on buildings makes it possible, for example to obtain interesting data on pollution levels in the air, energy consumption, and so on. In Singapore, the city has placed building-integrated sensors, which generate useful, reliable and instant data for the prevention and management of natural disasters (climate data, space data, photographs of urban areas, etc.).

The latest technological progress has helped to develop the generation, collection, storage, processing and dissemination of various forms of digital data at an unprecedented significant speed [HAM 07]. Moreover, it can be noted that coltan is one of the major sources of tantalum production (especially niobium) used for its high resistance to corrosion. This raw material is essential for making electrical components, especially surface wave capacitors and filters, used in the manufacturing of mobile phones and laptops.

Nowadays, we refer to Industry 4.0¹³ which tends to connect together all the means of production and makes their interaction in real time possible in an open space with no defined border. This is made possible thanks to NICTs, such as cloud computing, blockchain, augmented reality, big data analytics or the Internet of Things. This period sees the emergence of new actors such as the American giants GAFAM (Google, Apple, Facebook, Amazon and Microsoft), GAFAMITIS (Google, Apple, Facebook, Amazon, Microsoft, IBM, Twitter, Intel and Salesforce), or even the Chinese NATU (Netflix, Airbnb, Tesla and Uber) and BATX (Baidu, Alibaba, Tencent and Xiaomi) which represent the emblematic companies of the digital disruption¹⁴. According to a Gartner survey, by 2021, 20% of all human individual activities will be directly linked to at least one of the GAFA and BAT members. As an example, we can consider Apple with the launch in 2007 of the iPhone, which resulted in the creation of a market based on smartphones. On the other hand, China has developed big data in the maritime sector with the creation of a smart ocean. Therefore, Tsinghua University has set up a marine big data remote sensing center, taking into consideration the airwaves of oceanic storms, monitoring internal inundations and including an early warning system. Thus, Wu Lixin, the Director of the Qindao National Laboratory for Marine Science and Technology, proposed the project called Transparent Ocean, which is driven by satellite sensing, intelligent buoys, underwater gliders and robots, deep-sea space stations, and other various NICTs. The main challenge is to be able to, on the one hand, obtain overall information on the marine ecosystem at different