The new wave in journalism - robot writers?

So, how prevalent are robot reporters? What dangers do robot reports pose? What is the likelihood that the article you are reading wasn’t written by a human? A 2015 report stated the AP was generating about 3,000 articles a quarter, all related to financial and business reports. Some in the community have estimated that AI could generate up to 2,000 articles a second. In traditional, qualitative news, the Washington Post’s artificial intelligence technology, Heliograf, wrote 850 news stories in its first year (2016); most of these stories were related to the election. This trend points to more news articles being written by robots outside the financial field. But is this a good thing?

The Dangers of Robot Reporters

Robot reporters are great at compiling financial reports, and other articles related to numerical data (real estate, sports scores, etc.). However, they are not effective at compiling information from multiple sources with different opinions, tones, and goals. In fact, robot reporters are creating information overload, providing decision makers with huge quantities of often useless or even incorrect information. Several potential detriments to basing decisions on AI-written news are: 

 

Information overload: with so many articles being written due to automation, are the articles even relevant, or is there so much information available that it becomes meaningless

 

Useability: What is the importance of the information created by robot reporters considering that robots are good at translating statistics to words but aren’t great at synthesizing information to create something new?

 

Identifying authorship: does the programmer, the news outlet, or a human get credit for the article? 

 

Credibility: can software provide a truly unbiased and accurate article, or will it be overly-susceptible to bias in news sources?

Because of the sheer volume of information available and the uncertainty in the information’s source, it is more important than ever that managers curate their list of sources carefully. Relying on robot reports is dangerous and potentially costly, as it could cause managers to make decisions based on faulty information. Using tools that allow decision makers to quickly identify relevant data from massive amounts of information will provide managers with a competitive edge. These types of tools will be critical to the success of knowledge workers in the current news environment and in the future.

 

Given the wealth of benefits intelligent automation can bring, it is no surprise that news outlets are adopting AI to enhance their products. Machine Learning is a powerful tool for searching through multiple data sources and compiling important summary data for these sources. In particular, Natural Language Processing (NLP) provides a powerful tool for generating meaningful summaries of multiple text sources that highlights the most crucial information. Some NLP applications for information synthesis are found in other blog posts on this site. Some of the key tasks AI can perform for news reporting are:

  • Help expand audiences by increasing the number and topics of articles being produced. 
  • Support human reporters by automating basic fact compilation and synthesis. 
  • Monitor large sets of data, such as financial data or social media, and alert reporters on breaking trends. 
  • Provide templates and outlines for articles to save reporters time upfront. 
  • Reduce errors in data analyses.

Given these benefits, many news sources have deployed robot reporters. These are NLP algorithms that take numerous data source inputs and output a brief yet detailed news article. While the rise of AI authorship in news has caused concern among many journalists, these bots are designed to help journalists, not replace them. The goal of these tools is to reduce the amount of time reporters spend searching for trending topics and compiling and analyzing data. In total, the AP predicts robot reporters will free up 20% of human reporters’ time, allowing them to focus on quality over quantity.

topic modeling for product managers

AI-enabled research management system for market-intelligence.