Speaker: Dr André Curtis-Trudel, Lingnan University
Abstract:
Recent high-profile scientific achievements by machine learning (ML) and especially deep learning (DL) systems have reinvigorated interest in ML for automated scientific discovery (e.g., Wang et al. 2023). The hope, in rough outline, is that ML or DL might be used to identify promising novel events, phenomena, hypotheses, or even models or theories more efficiently than traditional, theory-driven approaches to discovery. This talk considers some of the more specific obstacles to automated, DL-driven discovery in frontier science, focusing on gravitational-wave astrophysics (GWA) as a representative case study. In the first part of the talk, we argue that prospects for DL-driven discovery in GWA remain uncertain. In the second part of the talk we argue for a shift in philosophical focus towards the ways DL can be used to augment or enhance existing methods, and the epistemological costs and benefits associated with these uses. We argue that the primary epistemological benefit of many such uses is to decrease opportunity costs associated with investigating puzzling or anomalous signals. For this reason, such uses are perhaps best understood as contributing to the pursuitworthiness of certain hypotheses, and we close the paper by contrasting this analysis with a recent proposal due to Duede (2023).
