Is it possible that in the near future, AI Daily could be staffed entirely by AI?
In May 2020, Microsoft announced that they would be cutting jobs in their MSN news production team, replacing them with AI to perform tasks such as identifying trending news stories, rewriting headlines, or adding photographs. A spokesperson said that the decision “can result in increased investment in some places and, from time-to-time, redeployment in others”.
Microsoft’s transition to “robot journalism” isn’t the first of its kind. In September 2017, Urbs Media announced the launch of “Reporters and Data and Robots” (RADAR), a project granted funding by Google; in partnership with the Press Association, where Artificial Intelligence was to be used to write around 30,000 stories a month for various regional newspapers and media in the United Kingdom. AI was designed to draw on the public data of government departments in order to generate the stories, with other AI being used to select appropriate images and videos for the stories. A team of five reporters was selected to curate and edit stories and pick the best data-sets to work with. Three months after the project launched, over 40,000 individual stories had been produced, some of these even making the front page of newspapers.
These developments pose great questions for the future of journalism, perhaps the most pertinent one being: will AI take over journalism entirely? Well, no, at least not for a while. Although a human might not do much of the actual writing of articles, humans would still be required to edit the articles, fact-check details, remove inappropriate or irrelevant content and improve the quality and readability of the writing. Additionally, some article formats are much harder to replicate using machines than others, such as opinion pieces or articles talking about personal experiences. These articles are more likely to require empathy or “common sense”, something difficult to replicate using existing technologies. However given the journalistic capabilities of AI demonstrated in these case studies, as well as the ability of AI to pump out a large quantity of articles in a much shorter space of time, there could be many job cuts in the media sector as media companies and newspapers look to cut costs while maintaining the same output; much fewer journalists are needed to simply curate articles than to create articles. Furthermore, machine learning algorithms could be deployed to improve the quality of output, so that over time, the articles generated by the AI could become more relevant and appropriate and increase in quality, further reducing the need for humans in the process.
For the time being, the poetic scenario of AI Daily eventually being staffed entirely by AI is unlikely. However if the technology continues to develop, AI could greatly change the way we operate, performing many of the tasks that we are currently performing such as sourcing, writing and adding media to the first drafts of articles. As the technology improves, the quantity of journalists required would decrease and eventually, AI may be doing most of the journalism for us.
Credit: Google News