Title: How AI is transforming seismology: From slow slip in Chile to petabyte-scale data mining
Abstract: Over the last decade, a range of new artificial intelligence (AI) tools have provided a step change for seismology. From faster earthquake early warning algorithms, to tenfold larger earthquake catalogs, and direct scientific insights: AI is accelerating scientific progress. In this talk, I will highlight the key contributions of AI to seismology, providing examples of different ways AI is used. How can we obtain faster early warning? How can dense earthquake catalogs reveal the dynamics of volcanic sequences and slow slip events in real time? How can AI help us understand earthquake rupture physics? At the same time, I will discuss the limitations of AI and how advanced approaches, such as combining simulations with observations, including physical constraints, or foundation models, might be able to address these. Finally, I will provide a perspective of how these approaches can be scaled to petabytes of data and how they could connect disciplines across the geosciences.
