Google is testing a new AI technology codenamed "Genesis" that can generate news articles, according to The New York Times. The tech giant has reportedly demonstrated the tool not just for The Times, but also for executives at The Washington Post and News Corp, which owns The Wall Street Journal. Based on reports from people who witnessed the pitch, Genesis can whip up copy from the data fed to it, whether it's current events or other types of information. Apparently, Google believes journalists could use it as some sort of an assistant to automate tasks and free them up for other things. Some of the people who saw the demonstration described it as "unsettling."

TL;DR Abstract - I built ZombieWriter, a Ruby gem that will enable users to generate news articles by aggregating paragraphs from other sources. It can use either machine learning algorithms (Latent Semantic Analysis and k-means clustering) or randomization to generate human-readable articles. In this article, I demonstrate how ZombieWrtier can use machine learning to create a Markdown file containing 17 human-readable articles. After finishing the demonstration and comparing the output to a randomization process, I then explain possible "future research" plans to improve the text generation process. I am not yet ready to claim that this technology is disruptive. Machine learning is hot (to put it mildly). The paradigm of using data instead of code to program machines has been applied to solve a variety of real-world problems.