The Main Challenges of AI Ethics: Historical Contextualization, Black-Boxing, Social Biases, Labor Invisibility

In the research described here, I argue that an adequate approach to AI ethics should include the four topics below. My aim is to answer the question of which are the necessary topics that someone should have under consideration in order to make an adequate approach to AI ethics. First, a critical history of AI, which focuses not on the technical differentiations between previous and following technologies, but on the social, economic, and political context in which artificial intelligence is designed, developed, and used. Second, an overview of the issues that most of the time are described as AI ethics, such as fairness, accountability, and transparency, in order to have the ability to understand what is missing from these approaches. A study on the black box of AI is necessary, not only from a technical perspective, but mainly from a perspective that is directly related to the political, social, and economic reasons that enforce and reinforce this black box, revealing, among others, the social relations, the hidden labor, and the “unintelligence” that are hidden under this black box. Third, an analysis of specific cases through critical approaches which take into account capitalism, with all the social, political, and economic relations that are connected with it. In this way, the emergence of biases, inequalities, and discriminations, becomes not a bag, but the substance of AI. Fourth, a study on the hidden labor of AI and the concerns regarding the future of work and AI. The study on hidden labor which is related with AI, is important in order, first, to criticize the intelligence and autonomy of AI systems, and second, to make visible the terrible working conditions of some workers, as a try to change them. The discussion regarding the future of work should not only contain discourses regarding the circular function of capitalism or vague ideas about ethical implementations of AI in the workplace. An adequate discussion should take into account the social, political, and economic relations of our society and ultimately challenge the current form of capitalism. I argue that all the above should be included in an adequate study of AI ethics.