Rajiv discusses the use of deep learning in predicting aftershocks following earthquakes, highlighting the importance of starting with simple models before diving into complex neural networks. Kyle questions the absence of basic baselines in the study, raising doubts about the true value of the deep learning approach in this context.
