Abstract
Ethics is ubiquitous in most domains, requiring much deliberation due to its philosophical nature. Varying views often lead to conflicting courses of action where ethical dilemmas become challenging to resolve. The major driving forces to make such a decision can be discretized and simplified to provide an indication of the most ethical course of action in a context. Given the parallel ubiquity of AI systems in these domains, it becomes increasingly imperative to work towards building inherently ethical AI that holds the ability to reason morally. This work proposes the use of knowledge representation and neurosymbolic techniques to develop resources for inherently ethical AI. It presents a three-phase framework towards bridging the path to moral machines: (a) Applied Ethics Ontology to make explicit the abstract concepts and relationships, (b) Dataset and graph generation using LLMs to develop a benchmark data store for ethical reasoning, and a (c) Case-based Reasoning algorithm to implement the philosophical concept of casuistry to make moral judgments and resolve ethical dilemmas.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call