Information technology has come a long way since its humble days as a hotline for journalists to ring when the computer said ‘No’.
When comedy writers poked fun at the stereotypical bored technician with the catchphrase, ‘have you turned it on and off again?’ could they have foreseen their meteoric rise?
Today’s digital landscape has not only radically transformed IT departments, it’s changed the way newsroom staff structures operate. Journalists now exist in a symbiosis with their IT counterparts who play a fundamental role in the dissemination of news.
The speed freakery of robot reporters
In 2019, the LSE’s Journalism AI project surveyed 71 news organisations in 32 countries and reported that artificial intelligence (machines and automation) is being used in the media in a multitude of ways.
At the Wall Street Journal, R&D chief and product lead for editorial tools, Alyssa Zeisler, has a refrain she returns to regularly: ‘Let’s help the computer do the what, so you (the reporter) can do the why.’ The paper’s machines act as robot reporters using data to create drafts of articles or produce news alerts. This enables the human reporters to do what they excel at, which is using their contacts, doing the analysis and conducting further investigation.
AI can also monitor social media for breaking news with a speed no human can match. Reuters’ News Tracer programme scans every tweet on Twitter in real time (around 500 million per day) to generate a summary of the items it rates as newsworthy. This summary is produced through the use of AI natural language processing techniques.
News Tracer has helped the agency break stories long before other outlets. In an interview with the Columbia Journalism Review, Reuters’ executive editor for data and innovation, Reg Chua, said, ‘This is a speed game, clearly, because that’s what the financial markets are looking for’.
Multi-skilling at a new level: the many talents of AI
AI can help to signal trends like changes in analysts’ ratings, unusually good or bad business performance or company insiders who may suddenly have sold stock. In this way, the computer is acting almost like a news editor, suggesting topics that merit consideration.
Machines can track the spread of fake news, evaluate the truth of statistical claims and carry out time-consuming tasks like turning video scripts into instant rough cuts for human review. These are all beneficial resources providing newsrooms with an enriched flow of information from which to compile their stories.
Like Reuters, Bloomberg uses AI for many of the mundane processing tasks involved in news gathering. Where once journalists would spend time transcribing recorded interviews, voice recognition technology can now do the job in seconds. In the place of one reporter sitting on a phone call listening into a briefing call with company chairs and CEOs, a computer can monitor several briefing calls at once and highlight when something significant happens.
Automatic translation is also deployed. Bloomberg reporters can break news in the language they’re most comfortable with and the computer can translate it. People are still required for accuracy checks but AI is helping improve the speed of news production and free up time.
Up close and personal: AI’s ability to tailor news
What looks like being its biggest impact on journalism, is AI’s capacity to personalise news for readers. Who, after all, has time to sift through information of no relevance as they hurtle through today’s digital world? AI and its algorithmic powers can help create newsletters or emails containing specific news items that will appeal directly to the individual reader without the need for them to complete laborious box-ticking exercises about likes and preferences.
At the Daily Telegraph, AI is enabling different subscription models to be presented to customers at the paywall stage. When a user accesses content that’s hidden behind a paywall, that paywall can be tailored to the individual’s specific interests which maximises the likelihood of them converting and taking out a full subscription.
The Telegraph also uses it for more effective allocation of its print newspaper. By using AI to analyse its 40,000 distribution points, the company can understand where the greatest need for its papers is and adjust the number of copies accordingly. Not only does this help save on overheads, it cuts waste which in turn has a positive impact on the environment.
The human upper hand: AI’s limitations in problem solving
While AI is excellent at delivering results on clear, well-defined tasks like scraping the net for certain words or market information, that’s because it’s working with solid data.
What it can’t yet do is put those results into context. If the problem is fuzzy or messy, as many human problems are, AI will be as helpful as a chocolate teapot.
The so-called journalist’s hunch for instance, responsible for so many scoops, is something the machine can’t imitate. It’s a uniquely human experience and here the algorithm falls down.
AI and the news: safeguarding the future when there’s no off switch
Despite its shortcomings, AI is on the rise, so much so that in 2020, the Guardian newspaper published an article written entirely by a robot. GPT-3 is an AI language generator computer more advanced than any of its predecessor language generators. The article prompted unease about how humans will supervise robot activity as these technologies become more sophisticated and widespread.
While fully automated journalism is limited for now, newsrooms will at some point have to develop the editorial skills of their people for effective oversight of the machines.
Journalism and ethics are supposed to exist side by side. Where newsrooms have failed to remember this in the past, there have been public and humiliating repercussions. Given artificial intelligence has no off switch, there’s an urgent need to ensure proper monitoring of its ethical side in the race for profits inherent in its expansion. Questions have been raised about how AI can be used ethically, for instance in the creation of personalised news.
For media companies to continue to benefit from AI, it’s essential that those at board level understand its risks and are well-versed on how and where it’s being used within their organisation. If ethics play just as much a role in its ascendancy, the potential for harm can be minimised.
There’s no doubt that AI is already empowering journalists in ways unimaginable even just a decade ago. And with public confidence in traditional media at an all-time low, this collaboration with the tech wizards is a good thing enabling better investigation, more holding power to account and the production of relevant news that people want to read. As long as there’s awareness of the pitfalls and strong safeguards in place, it is possible for journalism and AI to forge a formidable partnership helping create a well-informed public ready to face the challenges ahead.
In the interests of transparency and openness and so you know what our motive is, Powered by Coffee develops WordPress tools, software and other digital solutions for media publishers. We also happen to care deeply about the issues facing journalism in the 21st century and want to safeguard independent media and news through the design of better tech.