This document discusses machine learning and its empowerment through technological advancements. It describes different types of communication technologies from human to human, human to machine, and machine to machine. It then explains why machine learning has advanced recently due to big data, increased computing power through GPUs, and the development of tensor processing units (TPUs) by Google to accelerate neural networks. The document outlines the 1st and 2nd generation of TPUs and their improved processing capabilities. It concludes with a pitfall to avoid which is designing hardware using outdated models.