The Torch Lens Maker project is a significant advancement in the field of optics by leveraging PyTorch, a machine learning framework, to create a library focused on differentiable geometric optics. This library allows researchers and developers to obtain exact solutions to optical problems by treating optical elements as layers in a neural network. This innovative approach mirrors the architecture of neural networks where each layer manipulates input data—in this case, rays of light—based on their specific parameters, leading to an efficient method for optimizing optical systems. There is a possibility of expanding this work to include diffractive optics, drawing interest from those experienced in holography and its visualization techniques.