Dr James Edgar Lim1
1National University Of Singapore, Singapore
Biography:
James is a research fellow at the National University of Singapore's Centre for Biomedical Ethics. He wrote his PhD thesis at the Australian National University, on the Ethics of Public Shaming. He was appointed a Associate Lecturer at the ANU, before moving on to the CBmE in July 2024.
James’ research interests include issues in applied ethics and political philosophy, including AI ethics, bioethics, and issues related to social punishment.
Abstract:
"Trustworthy AI" is a buzzword in AI research, with scores of scholars discussing whether AI systems are trustworthy, and how developers can create trustworthy AI.
This demand for trustworthy AI is intuitive, especially in healthcare where the stakes are incredibly high. But philosophers usually make a distinction between trustworthiness and reliability. Trustworthiness is defined as reliability, in addition to some other trait, such as virtue, goodwill, or reasons-responsiveness (for example, your mug might be reliable, but not trustworthy in the sense that people are).
These philosophical observations leave us with the question: do we need trustworthy AI, or merely reliable AI? In this presentation, I argue that all we need (at least in the healthcare context) is reliable AI. The goods of trust – the value that trust and trustworthiness bring to our lives – include: the expression of respect (for the trusted person); the promotion of cooperation; the development of meaningful relationships; the exercise of autonomy; and the promotion of moral maturity. These are primarily goods that are enjoyed within the context of relationships between humans. These are not goods that can be enjoyed in the context of human-AI relationships. Thus, we do not gain any distinct value from trusting AI systems, or considering AI systems trustworthy. Rather, it makes more sense to think about how the institutions that use AI systems can be trustworthy. In the context of healthcare, what's important is whether healthcare practitioners and institutions are trustworthy.