Minimizing Action with Gradient Descent: A Novel Physics Perspective

This post presents a unique perspective on physics: viewing it as an optimization problem. The author solves the free-fall problem by minimizing the action using gradient descent, instead of traditional analytical or numerical methods. The post compares analytical, numerical, and action-minimization approaches, implementing the latter with PyTorch. The results match analytical and numerical solutions, offering a fresh perspective on classical mechanics and paving the way for exploring more complex physical systems.