Societal Responsibility of Artificial Intelligence

Оглавление

Группа авторов. Societal Responsibility of Artificial Intelligence

Table of Contents

List of Illustrations

List of Tables

Guide

Pages

Societal Responsibility of Artificial Intelligence. Towards an Ethical and Eco-responsible AI

Acknowledgments

Introduction

ALGORITHMIC PROCESSING.–

BIG DATA.–

BLOCKCHAIN.–

CROWDSOURCING.–

DEEP LEARNING.–

WEAK AI.–

STRONG AI.–

MACHINE LEARNING.–

ETHICS BY DESIGN.–

ETHICS BY EVOLUTION.–

1. Societal and Moral Questioning Around AI and Its Ecosystem

1.1. Use cases of AI

CHATBOTS.–

1.2. Digital environment

DATA DRIVEN.–

API.–

CLOUD COMPUTING.–

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

BUSINESS INTELLIGENCE.-

1.4. Technological and societal issues

ADMINISTERED DATABASE.–

BIG TECH.–

FREE AND INFORMED CONSENT.–

CYBERSECURITY.–

1.5. Ethical and moral issues

OPEN SOURCE.–

ANONYMIZATION.–

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

OPT-OUT.–

Box 2.1.Biases in numerical data

Box 2.2.Effects of algorithms on individuals

DYNAMIC PRICING.–

FAKE NEWS.–

Box 2.3.Learning algorithms and transparency

COMMUNITY MANAGER.–

ACCOUNTABILITY.–

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

TURING.–

MERISE METHOD.–

SOURCE CODE.–

2.4.1. Data ethics

2.4.2. Ethics of algorithms

2.4.3. Ethics of systems

2.4.4. Ethics of practices

2.4.5. Ethics of decisions

2.5. Analysis of AI-related knowledge

CYBERNETICS.–

COMPUTOSPHERE.–

INFOSPHERE.–

3. Ethical Framework Associated with AI

INDUSTRY 4.0.–

LIVING LABS.–

3.1. Ethical charter around AI

3.1.1. Data ethics

3.1.2. Ethics of systems

3.1.3. Ethics of algorithms

3.1.4. Ethics of practices

3.1.5. Ethics of decisions

PRIVACY BY DESIGN.–

3.2. Recommendations for AI

3.2.1. Data ethics

3.2.2. Ethics of systems

3.2.3. Ethics of algorithms

3.2.4. Ethics of practices

3.2.5. Ethics of decisions

OPEN DATA.–

SANDBOX.–

NOSQL.–

HADOOP.–

Box 3.1.Ten key points of ethical data

GDPR.–

DATA SCIENTIST.–

DATA ENGINEER.–

DATA STEWARD.–

DATA ANALYST (BUSINESS ANALYST).–

DATA OWNER.–

3.3. Temporality relative to the human guarantee in digital technology

HUMAN GUARANTEE OF AI.–

3.4. For the health user and for health user representation

EMPOWERMENT.–

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

TEST AND LEARN.–

ROI.–

3.6. Environmental parameters of digital technology

DEEP NEURAL NETWORKS.–

Box 3.2.Environmental parameters of NICTs within a company

3.7. Regulation associated with AI

3.7.1. Structural parameter

3.7.2. Technological parameter

3.7.3. Strategic parameter

3.7.4. Methodological parameter

3.7.5. Organizational parameter

3.7.6. Regulatory parameter

3.7.7. Relational parameter

3.7.8. Cultural parameter

DATA VISUALIZATION.–

3.8. Algorithmic systems and digital data governance

DATA MINING.–

PUSH.–

COOKIES.–

LOG.–

3.9. Four key steps for an AI project. 3.9.1. Step 1: determine the project objective

3.9.2. Step 2: collect and prepare relevant data

KNOWLEDGE GRAPH.–

3.9.3. Step 3: classify data and choose tools

3.9.4. Step 4: produce the model

3.10. Algorithmic responsibility

4. Anticipation Around Artificial Consciousness

4.1. Protean aspects of consciousness associated with intelligence

Box 4.1.Different forms of intelligence

4.2. Structuring of consciousness

Box 4.2.Structuring consciousness

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

Box 4.3.Knowledge communication processes

4.4. Process of creating practical wisdom from artificial consciousness

4.5. Morality of a “strong” AI

Box 4.4.Pitfalls to avoid for “strong” AI

Conclusion

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

A7.1.1. Humanity and ethics

A7.1.2. Data science, the art of artificial intelligence, privacy and personal data

A7.1.3. Daily work and etiquette

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. A, B

C

D, E

F, H, I, J

M, N

P

Q, R

S

T, U

WILEY END USER LICENSE AGREEMENT

Отрывок из книги

Technological Prospects and Social Applications Set

.....

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.

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).

.....