Knowledge representation and acquisition for ethical AI: challenges and opportunities

Abstract

Machine learning (ML) techniques have become pervasive across a range of different applications, and are now widely used in areas as disparate as recidivism prediction, consumer credit-risk analysis, and insurance pricing. Likewise, in the physical world, ML models are critical components in autonomous agents such as robotic surgeons and self-driving cars. Among the many ethical dimensions that arise in the use of ML technology in such applications, analyzing morally permissible actions is both immediate and profound. For example, there is the potential for learned algorithms to become biased against certain groups. More generally, in so much that the decisions of ML models impact society, both virtually (e.g., denying a loan) and physically (e.g., driving into a pedestrian), notions of accountability, blame and responsibility need to be carefully considered. In this article, we advocate for a two-pronged approach ethical decision-making enabled using rich models of autonomous agency: on the one hand, we need to draw on philosophical notions of such as beliefs, causes, effects and intentions, and look to formalise them, as attempted by the knowledge representation community, but on the other, from a computational perspective, such theories need to also address the problems of tractable reasoning and (probabilistic) knowledge acquisition. As a concrete instance of this tradeoff, we report on a few preliminary results that apply (propositional) tractable probabilistic models to problems in fair ML and automated reasoning of moral principles. Such models are compilation targets for certain types of knowledge representation languages, and can effectively reason in service some computational tasks. They can also be learned from data. Concretely, current evidence suggests that they are attractive structures for jointly addressing three fundamental challenges: reasoning about possible worlds + tractable computation + knowledge acquisition. Thus, these seems like a good starting point for modelling reasoning robots as part of the larger ecosystem where accountability and responsibility is understood more broadly.

Navigating Applied Artificial Intelligence (AI) in the Digital Era: How Smart Buildings and Smart Cities Become the Key to Sustainability

“…The literature review helped to summarize existing research in closely related fields of interest and providing a conceptual framework that facilitates direct future work toward deepening research (Torraco, 2016). Having increasing access to continuously increasing available quantity of experience and data sets from 2021 to date enables quantitative studies and case studies as part of the state-of-the-art (Abdirad & Mathur, 2021;Belle, 2023;Prabhakar et al, 2023).…”

Application of machine learning in tax prediction: A review with practical approaches

“…Recent studies have delved into the ethical complexity of AI and ML, encompassing issues such as bias, justice, responsibility, and autonomous decision-making (Belle, 2023;Gordon, 2021;Dan, 2018). In the realm of taxation, these concerns are particularly pertinent, highlighting the necessity for responsible usage of ML models.…”

“…In the realm of taxation, these concerns are particularly pertinent, highlighting the necessity for responsible usage of ML models. This involves a comprehensive understanding of their limitations and the potential impacts they may have on taxpayers (Belle, 2023).…”

Concerns about the role of artificial intelligence in journalism, and media manipulation

“…It has become a field of study, finding its way into academic research and taught university programmes; within the (UK) academic sphere there seems to be a current focus on machine-learning, language processing, education, and knowledge representation. 1 A brief literature search indicates the flexibility and diverse range of academic research that makes use of AI as a topic as well as those utilising it as part of their toolkit (Belle, 2023). Indeed, the term most widely used is simply AI, with the acronym seemingly divorced from the words it represents.…”