Editorial decisions on what stories news services publish could soon be made by artificial intelligence instead of human professionals. Is this the end of the publishing industry as we know it, or just another innovation to attract readers?
Imagine: No more editors deciding what to write about, looking at angles, assigning stories to journalists and driving wider conversation with what’s included on the front page of a publication. All of that is decided by artificial intelligence (AI), a machine.
It’s not a reality yet, but this is the scenario that could develop if Entire Digital Publishing’s business model proves successful.
The Milan-based start-up began publishing its flagship online newspaper notizie.it less than two years ago.
The publication, which uses an algorithm to inform what stories are commissioned and published, is now among the most viewed news sites in Italy while a Spanish language version of it (for Spain and Latin American nations) is on the cards.
“We realised that online publishing and journalism world had not evolved in the past 20 years, since the late 90s when the digital revolution occurred,” Entire Digital Publishing’s CEO Massimiliano Squillace tells SBS Italian.
According to Squillace, at that time publications moved their content from physical platforms to digital, but maintained the internal structures, roles and workflows of traditional print media organisations.
“I spent a month at the Washington Post – one of the best online newspapers in the world – and I realised that, in spite of being owned by Amazon’s Jeff Bezos, they have maintained a rather traditional corporate organisation,” says Squillace.
Entire Digital Publishing is trying to operate in a less traditional way.
“Instead of giving readers the news, we understand through an algorithm what readers want to know and read,” he says. “Then we write those stories for them and make them available to them.”
The rise of ‘robot-journalists’?
“Journalists’ work is sacred to us,” said Squillace, adding that he does not believe in US experiments with news-writing “robot-journalists”, in spite of the millions invested.
Therefore, people still write the articles for notizie.it, but who writes what is decided by the machine.
“Our AI assigns the right job to the right person choosing from a pool of 10,000 freelance journalists,” he said. The decision is based on criteria such as the writer’s interests, expertise, track record and location.
But before finding the right person for the job, the AI searches for the best article topics and angles.
“One of the main features of our algorithm is content discovery,” he says. “We are developing it in such a way that it will be able to scan the web to find angles and stories that are trending.”
And readers themselves can become the actual sources of news.
“For example, if an incident happens in a certain neighborhood, something about it is normally shared by local residents on public groups or social media before the story gets covered by news outlets,” he says.
Massimiliano Squillace explained that he and his team started trawling social media for news manually, initially scanning groups of neighborhoods in Milan, looking for publicly-posted news.
“For example, if someone posted something like ‘there was an explosion’, we fact-checked it and then produced a news item. That sort of news was very successful in terms of views and engagement.”
Another aspect of the AI employed by Entire Digital Publishing is content recommendation.
“We suggest not only articles consistent with your interests but also new content that we think you could be interested in and you would have never considered, thus bursting readers’ ‘cognitive bubble’.”
Beware the machines?
Is this another example of new technologies stealing traditional jobs? The answer may not be that simple.
“We often tend to see technology as an alternative to humans while it is rather a complementary factor,” said Max Tani, Professor of Finance at the University of New South Wales in Canberra.
According to Tani, automated processes can provide businesses with a competitive advantage – as in the case of Entire Digital Publishing – but a human editor will nonetheless be required to verify and ultimately make certain decisions in particular occasions and extreme circumstances.
It would appear that for now, humans and machine can still peacefully coexist.