Machine learning, which draws from a diversity of fields including computer science, mathematics, and physics, has been taking the world by storm due to its flourishing industrial applications. We present a live demo of machine learning in action, where we train a neural network to classify handwritten digits to an appreciable degree in real time. We then proceed to give an introductory overview of the ingredients that go into this task: logistic regression, stochastic gradient descent, and deep learning. (Part II, which connects machine learning to approximation theory and quantum physics, will take place the following week.)