Embedding Ethics Into Artificial Intelligence: Understanding What Can Be Done, What Can't, and What Is Done

Abstract

Embedding ethical considerations within the development of AI driven technologies becomes more and more pressing as new technologies are developed. Given the impact of autonomous technologies on individuals and society, it is worth taking the time to assess and manage the ethical aspects and possible consequences of our technological endeavors. While the growing rapidity of autonomous decision processes makes it hard to keep individuals in the decision loops, people are turning their attention to the ways in which ethics could be integrated to machines and algorithms, as well as to the possibility of defining autonomous ethical machines that would be able to solve ethical dilemmas and act ethically (e.g. autonomous vehicles). Notwithstanding theoretical and practical difficulties surrounding the possibility of defining such ethical machines, important elements should be considered when reflecting on the embedding of ethics into AI technologies. The present paper aims to critically analyze the limitations of such endeavors by exposing common misconceptions relating to AI ethics.