Data journalism chiefs from CNN, the BBC, Bloomberg News and United Robots have told how advancements in artificial intelligence have supported the growth of reporting teams rather than threatened jobs.
In a discussion at Press Gazette’s Future of Media Technology conference in London, panellists were asked to confront fears about automation leading to fewer newsroom jobs.
“I get this question all the time, both internally and externally,” said Claudia Quinonez, managing editor for news automation at Bloomberg News. “The interesting thing is that it has not materialised.
“So we’ve been doing this for eight years, and it’s just not what happens. You do not automate people out of their jobs. You actually automate tests that they hate doing in a very precise way.”
She spoke about how, in the past, journalists would need to start scheduling articles to go live online in the early hours “so you can get a piece at 7am when your readers want to read. The bot can do it super-easy – it’s ready for you, served on a platter, and you just pull it up.”
“So really, at the end, we’re best if we look at technology as our friend. Because it can help us do incredible journalism. It can help us find incredible data trends which I think enables… the group of journalists [with] sound news judgement and a knowledge of which trends are really relevant. And we just help them do that faster and better.”
Deepna Devkar, vice president of data science and engineering at CNN, said: “It’s actually the exact opposite. We’re here to enable journalists to do what they do best…
“We’re not at all here to automate journalism. I think creativity will never be replaced by machines, and it can’t be. That’s what makes us us. And so, we have a very strong partnership with our editorial team.”
Devkar’s CNN team uses data tools to assist newsrooms with tasks that are “impossible to do it with a human eye”.
“So really, at a very high level, we do things that machines can do better than us. Which is pattern recognition among big amounts of information that humans just can’t parse through that quickly. Building a knowledge graph for content that basically gives them access to what are the topic clusters that naturally fit together, how are they related in a meaningful way, those kinds of things. And then finally personalisation, which is we’re doing a lot of work on user personalisation while respecting the user privacy…
“So those are the things that journalists can’t do themselves and so those are the things that we enable them to do.”
John Walton, data journalism editor at the BBC, explained how his team uses data tools to “automate repeatable tasks”.
“The way automation works for us is that it widens the net of things we can do,” he said. “Another good example of the way we automate content is election results. When election results are coming in for local councils or local constituencies, we can tweet out a graphic of each single result as they come in. It’s more or less instant, from the data feed to the publication of the graphic and the caption on Twitter, the text linking back to the site.
“So that is something where you might be producing several hundred graphics a night, overnight during an election result, and you just would not be able to do that manually.”
Cecilia Campbell, the chief marketing officer of United Robots, a Swedish tech firm that supplies local news publishers with automated content, said her company’s tools help journalists focus on more traditional reporting.
“For us, it’s very much about content automation – we use the robots to do the routine reporting,” she said. “That does two things. It frees up journalists, that you don’t have that many of anyway sometimes in some of these newsrooms. And it also means you can do so much more – you can do so many more stories, or create that much more information.
“I would say I think it’s a complementary thing. Those incremental stories may not be great journalism. But again, combined with the bigger stories of the day, you generate something that’s really valuable to local people.”
She told of one Swedish local publisher that runs daily stories across its sites from United Robots. “And in parallel to deploying these robots, that group has actually filled 25 new journalism jobs as well. Because it’s all about investing in their product – local journalism.”
Photo credit: ASV Photography: Pictured (left to right) are William Turvill, Cecilia Campbell, John Walton, Deepna Devkar and Claudia Quinonez
Email pged@pressgazette.co.uk to point out mistakes, provide story tips or send in a letter for publication on our "Letters Page" blog
