Ethical AI Enchantment: A Delusion or Dilemma?

Abstract

The rapid rise of Artificial Intelligence (AI) has introduced a dual narrative, one can regard it as a novel of transitional metamorphosis and another as a novel of moral concern. While people tirelessly call for new advances in AI and its capability of atomizing industries and such systems, this belief tends to ignore the actual potential gap widening between the beliefs held regarding AI and the effective performances of AI that does exist. This paper explores this gap critically noting that, inflated expectations of what AI brings to the table often do not take into consideration its virtues and vices such as, bias, invasion of privacy, tampering with sheer autonomy and obliviousness to responsibility. Today’s AI systems, which rely on large datasets, tend to replicate historical disparities and contribute to prejudice in convergent results while giving voice to systemic unfairness. Moreover, due to increasing AI decision making power, its internal mechanisms, frequently described as ‘black boxes’, create major issues regarding responsibility attribution and, especially, transparency. Such limitations indicate the relevance of ethical issues that surround the application of artificial intelligence, especially in areas that are closely related to our lives such as health, finance, and justice systems. This paper shows that although ethical supervision, parity, and diversity are important, present technological growths still lack the capacity to handle the complexities of such issues effectively. To address this gap, it is necessary to advance in technology as well as to coordinate the close collaboration between policymakers, developers, and society members. Systems ethics alone thus are not sufficient: it is only through a strong ethical programme based in particular on the principles of openness and responsibility that the Cognitive Intervention sought by AI can reconcile its potential with its values.