Over the last few decades, we’ve seen journalism evolve in a number of ways, moving with the times through print, radio, broadcast, and now digital journalism mediums. Today, we see the rise of a new form using Artificial Intelligence to create automated journalism. Using Natural Language Generation (NLG), automated data journalism is able to take data collected and turn it into news stories without a journalist having to come in contact with the story. Columbia University’s “Guide to Automated Journalism” explains that it “is most useful in generating routine news stories for repetitive topics for which clean, accurate, and structured data are available,” which fulfills “news organizations’ aim to both cut costs and increase the quantity of news.”
While it can seem like a negative update in the face of what seems to be a dwindling industry, it does allow for journalists to focus on the more important stories, furthering the process a little quicker to enable reporters to jump onto a case once it’s deemed relevant enough. The little things, however, can be easily taken care of through AI. Additionally, redundant and tedious work, such as data, is easily avoidable, and it takes out the chance of human error, providing more accurate and detailed reporting. Programs also are able to filter out fake news, ensuring that the stories are as factual as possible, which is important when readers aren’t as quick to fact check with media sources. There is also an increasing demand to facilitate more communication between readers and reporters, which robotic journalism accomplishes, as it can respond to comments and requests for more information in real time, even managing the content to hide offensive comments. Overall, this recent development is sure to grow, as it provides the industry with more accurate, up-to-date, and effortless reporting.
Quite a few news organizations have been working on automated journalism, incorporating it into their readers’ experience in quite a few different ways. One of these examples is Reuters, which uses robo journalism to create interactive data graphics, updated in real time, using a semantic technology company called Graphiq. Through data visualization, visitors to their site can have data explained to them in an easily consumable manner, while ensuring that their numbers are up-to-date and accurate. Additionally, it saves the journalist the time of altering the graphic every time new information is available.
Automated Insights works with Yahoo!, who mainly uses robot journalism for their sports and finance sections. They provide the following video detailing their experience with Yahoo! Sports:
Ever since the big year of 2016 with the Summer Olympics and a majorly controversial presidential election, The Washington Post has used the automated journalism system Heliograf, even winning an award for excellent use of bots in the 2019 Global Biggies Awards, an award ceremony focusing on AI and data. Their methods go so far as to automate certain stories for specific regions using geo-targeting, ensuring that the right information reaches the best audience.
The New York Post has been using AI since the release of their AI software Editor, which allows for quicker writing in the form of research and fact checking, searching through relevant information using simple semantic tags. More recently, they have implemented programming called Perspective API, created by Jigsaw, acting as a type of moderator through their comment sections by judging comments based on their toxicity levels. This allows readers to filter through messages without seeing the harsher phrasing that others may have left through a handy sliding scale.
You can hear more about Jigsaw’s mission here:
BBC also uses semantic technology in order to group information together in more accessible manners. By picking out keywords involving topics, journalists looking to conduct research simply have to select a few terms to narrow down the search, and they can be easily pointed in the right direction for quick and painless research. Juicer, the program they use to extract this data, explains this process:
Bloomberg Media is known for having a third of all their content produced by robo journalism, using a program called Cyborg in particular, mainly for financial reporting of earnings reports. Forbes has also joined in, their program being Bertie, which provides their reporters with templates for news stories and even rough drafts, cutting out a lot of the work.
By pulling data and using pre-written templates, the Los Angeles Times has also been using AI, specifically in covering earthquakes and homicides. As there seems to be an abundance of both, journalists are then able to use robotic data journalism in order to increase quantity and focus their efforts on reports that take more work.
All the way back in 2014, Associated Press began using Automated Insights, which is an NLG software that allows them to turn out data journalism at a much faster rate. In fact, while they used to produce 300 articles on earning reports each quarter, they now have 3,700. NewsWhip shows this example of their syndication that tracks social engagement with AP’s content:
Further still, other news organizations, such as The Guardian and Quartz Digital News, have employed AI to run chatbots, some of which allow visitors to type in questions and receive relevant information or even get their news summed up in a simple manner, delivered to their inboxes at a certain time each day. These are excellent methods to provide more interaction between readers and the organizations, all without having an actual journalist needing to push a button.
Through all of these examples, one idea rings clear: automation in journalism, or even just the business of content creation, allows for a greater focus on the parts that really matter. Allowing AI to take care of the minor things is a gateway to being able to avoid pouring over tedious data and instead channel creativity and insight into bigger topics. As creatives, this is an excellent way to improve quantity, as well as the quality, by allowing creatives to do what they do best, all while providing readers with an easy to understand experience.
