Explore the application of Physics-Informed Neural Networks (PINNs) to ice and climate modeling in this 23-minute conference talk. Learn how machine learning techniques can be integrated with physical principles to better understand and predict polar climate systems. Discover the PINNICLE framework and its potential for advancing climate science through the combination of neural networks with established physics equations. Examine how this innovative approach addresses challenges in modeling complex ice dynamics and climate interactions. Gain insights into cutting-edge computational methods that could improve projections of polar climate evolution in a warming world, presented as part of a comprehensive conference on polar climate system understanding and future predictions.