Google Testing AI Tool to Automate News Article Creation

According to The New York Times, Google is currently testing a new AI technology called “Genesis” that has the ability to generate news articles. The company has reportedly showcased this tool to executives from The Washington Post and News Corp, who own The Wall Street Journal, in addition to The Times. Witnesses of the demonstration claim that Genesis can quickly produce written content based on the data it is provided, whether it pertains to current events or other forms of information. Google envisions journalists utilizing this technology as an assistant to automate certain tasks, allowing them to focus on other responsibilities.

Some people who witnessed the demonstration described it as “restless”. They also said it seemed to ignore the work that goes into writing precise, fluid songs. Jeff Jarvis, a professor of journalism at the City University of New York, told The Times that journalists should use the tool “if this technology can reliably deliver factual information.” Since Google hasn’t released Genesis yet, we can’t say if it actually can, or if it could easily lead to the spread of misinformation. As The Times points out, Google has moved quickly to adopt AI technology in an effort to catch up with Microsoft-backed OpenAI. Its generative AI technology, Bard, was caught spreading misinformation as soon as it debuted on Twitter.

Recent attempts by some publications to use AI tools had not ended well. CNET was forced to correct it after it became aware of significant errors in most of the 77 typed articles it published on the CNET Money page. And just earlier this month, Gizmodo’s io9 published a Star Wars piece full of errors attributed to the “Gizmodo bot.” James Whitbrook, io9’s deputy editor, said the website’s editorial team had no part in publishing it, so it had no chance to edit it for corrections before it was published.