[LINK] Data-Driven Understanding of Real-Life Moral Dilemmas

[Tom Worthington]

Title: Data-Driven Understanding of Real-Life Moral Dilemmas
Speaker: Josh Nguyen (The Australian National University)
When: Thursday, 20 April 2023, 11am - noon
Where: ANU Building 145, Room 1.33 and zoom 
https://anu.zoom.us/j/9874074503?pwd=bzlETnF1VEUzcyswZDBMTktQRUYyZz09


Abstract

The emergence of AI systems—trained on massive datasets and seemingly 
capable of engaging in moral discourse—raises several important 
questions about automated moral decision making. While the philosophical 
literature largely focuses on idealized moral dilemmas, we aim to 
examine the patterns of human moral judgment and sentiment observable in 
large-scale social media datasets, which can provide insight into 
real-world ethical issues. In this MPhil completion talk, we will 
present two such studies. First, we explore the features of everyday 
moral conflicts through a topic-driven analysis of over 100,000 moral 
discussion threads on Reddit. Our results suggest that many complex 
moral dilemmas arise from nominally neutral topics such as money, work 
and appearance. Moreover, these topics serve as an important covariate 
in examining the ways in which a moral story is framed and how a 
judgment is made. In the second study, we operationalize the moral 
foundations theory—which proposes five fundamental dimensions of 
morality—in analyzing content on Reddit and Twitter. Consistent with 
prior research, we find that the relative importance of these moral 
dimensions can explain the variance in people's stance on several 
socially relevant topics like atheism and the legalization of abortion. 
We also find that opposing judgments in the same moral conflict can be 
characterized by the moral dimensions they underscore. Together, these 
studies demonstrate the utility of a data-driven approach to practical 
ethics, especially given the rising presence of AI-assisted solutions.


Bio
Josh Nguyen is an MPhil student at the Computational Media Lab, 
supervised by Lexing Xie (School of Computing) and Colin Klein (School 
of Philosophy). He is also affiliated with the Humanising Machine 
Intelligence Grand Challenge at ANU. His broad interests include 
computational social science, machine learning and optimization. He 
received his BSc in computer science from the University of Melbourne in 
2020.



-- 
Tom Worthington http://www.tomw.net.au