Concepts of Ethics and Their Application to AI

Any discussion of the ethics of AI needs to be based on a sound understanding of the concept of ethics. This chapter therefore provides a brief overview of some of the key approaches to ethics with a particular emphasis on virtue ethics and the idea of human flourishing. The chapter reviews the purposes for which AI can be used, as these have a bearing on an ethical evaluation. Three main purposes are distinguished: AI forefficiency, optimisation and profit maximisation, AI forsocial control and AI for human flourishing. Given the focus on human flourishing in this book, several theoretical positions are introduced that provide insights into different aspects and ways of promoting human flourishing. The chapter concludes with a discussion of the currently widespread principle-based approach to AI ethics.

AI Ecosystems for Human Flourishing: The Recommendations

This chapter develops the conclusions that can be drawn from the application of the ecosystemmetaphor to AI. It highlights the challenges that arise for the ethical governance of AI ecosystems. These provide the basis for the definition of requirements that successful governance interventions have to fulfil. Three main requirements become apparent: the need for a clear delimitation of the boundaries of the ecosystem in question, the provision and maintenance of knowledge and capacities within the ecosystem, and the need for adaptable, flexible and careful governance structures that are capable of reacting to environmental changes. Based on these requirements, the chapter then spells out some recommendations for interventions that are likely to be able to shape AI ecosystems in ways that are conducive to human flourishing.

Addressing Ethical Issues in AI

This chapter reviews the proposals that have been put forward to address ethical issues of AI. It divides them into policy-level proposals, organisational responses and guidance for individuals. It discusses how these mitigation options are reflected in the case studies exemplifying the social reality of AI ethics. The chapter concludes with an overview of the stakeholder groups affected by AI, many of whom play a role in implementing the mitigation strategies and addressing ethical issues in AI.

Introduction

The introductory chapter describes the motivation behind this book and provides a brief outline of the main argument. The book offers a novel categorisation of artificial intelligence that lends itself to a classification of ethical and human rights issues raised by AI technologies. It offers an ethical approach based on the concept of human flourishing. Following a review of currently discussed ways of addressing and mitigating ethical issues, the book analyses the metaphor of AI ecosystems. Taking the ecosystems metaphor seriously allows the identification of requirements that mitigation measures need to fulfil. On the basis of these requirements the book offers a set of recommendations that allow AI ecosystems to be shaped in ways that promote human flourishing.

Conclusion

The conclusion briefly summarises the main arguments of the book. It focuses on the requirements for mitigation options to be used to address the ethical and human rights concerns of artificial intelligence. It also provides a high-level overview of the main recommendations brought forth in the book. It thereby shows how conceptual and empirical insights into the nature of AI, the ethical issues thus raised and the mitigation strategies currently being discussed can be used to develop practically relevant conclusions. These conclusions and recommendations help to ensure that AI ecosystems are developed and shaped in ways that are conducive to human flourishing.

Ethical Issues of AI

This chapter discusses the ethical issues that are raised by the development, deployment and use of AI. It starts with a review of the (ethical) benefits of AI and then presents the findings of the SHERPA project, which used case studies and a Delphi study to identify what people perceived to be ethical issues. These are discussed using the categorisation of AI technologies introduced earlier. Detailed accounts are given of ethical issues arising from machine learning, from artificial general intelligence and from broader socio-technical systems that incorporate AI.

AI Ecosystems for Human Flourishing: The Background

This chapter analyses the concept of AI ecosystems with a view to identifying how the ecosystem metaphor can help deal with ethical questions. The first step is to introduce in more detail the concept of ecosystems, drawing specifically on the literature on innovation ecosystems. This allows the identification of characteristics of ecosystems such as their openness, the co-evolution and mutual learning of their members, and the interdependence and complex relationship between those members. These characteristics underlie the challenges that an ethics-driven approach to ecosystems must consider.

Perspectives on Artificial Intelligence

A discussion of the ethics of artificial intelligence hinges on the definition of the term. In this chapter I propose three interrelated but distinct concepts of AI, which raise different types of ethical issues. The first concept of AI is that of machine learning, which is often seen as an example of “narrow” AI. The second concept is that of artificial general intelligence standing for the attempt to replicate human capabilities. Finally, I suggest that the term AI is often used to denote converging socio-technical systems. Each of these three concepts of AI has different properties and characteristics that give rise to different types of ethical concerns.