Getting your Trinity Audio player ready...
|
With News Corp Australia producing a mind bending 3,000 articles a week using generative AI – using just a team of four – the question now is what is the real role of editorial teams? In a wide ranging interview, the doyen of AI, Professor Charlie Beckett, tells us that media companies will soon think of themselves as data companies, not newsrooms. For journalists, it will become solely about adding the human touch.
“I welcome our robot overlords in the sense that it means we will do less boring crap,” says media expert Professor Charlie Beckett as he contemplates how the rapid rise of artificial intelligence will transform editorial teams.
As the director of JournalismAI, an initiative by the London School of Economics’ Polis think-tank and the Google News Initiative, Beckett has been exploring ways for news organisations to use AI responsibly. And he predicts new tech will allow swamped content creators – already expected to be more multi-skilled than ever before – to get out of the office.
“Journalists increasingly have to be able to do everything and differently – be a curator, a filter, community activist, engagement specialist, and be revenue conscious,” Beckett points out. “AI adds to that, but also facilitates all that. It will make it possible for the journalist to be that multifarious thing.
I call it the jetpack journalist. The robot won’t replace you but you will have all these little robots working for you. When an Editor says do me a long piece, a short piece, a snippet for Twitter, a piece for TikTok, etc, you’ll press an F key and the AI will reformat it all for you.
Professor Charlie Beckett, London School of Economics & Political Science
“And while becoming this jetpack journalist sounds great, it’s not easy. You may get distracted and spend all your time managing this increased capability instead of doing what you should be doing, which is adding human value – talking to people and going out of the office. AI will hopefully enable you do this by getting across those feeds you have, those newsletters and emails you get every morning and summarising them for you.”
Finding time to connect with the public will allow journalists to add some much-needed personality to editorial teams driven by AI. “If a machine can do the bland, easy news, why should someone go to your news website and not just use ChatGPT?”
Your company’s USP is going to be – we have empathy and go out there and work with our community and witness what is happening. A machine can’t do that.
Professor Charlie Beckett, London School of Economics & Political Science
The importance of data and research
The rise of AI will see editorial teams become far less siloed and much more strategic, according to Beckett. In particular, media groups will be more focused on the data they bring in.
“They’re going to think of themselves as data companies rather than as newsrooms,” says Beckett. “The nucleus will always be the editorial team, but how you generate content will be so important, especially the use of audience data. The most important thing is that we know what our audience does.”
Over the last few months, some news organisations have played around with ChatGPT, getting the chatbot to write articles. According to Beckett, experimenting with AI should be done more systematically.
“These large language model based tools weren’t meant to write a breaking news article, so why are you doing that? You wouldn’t use a spade to cook a dinner. The best organisations are experimenting by setting up a working group who are deliberately, systematically working through potential use cases, thinking – which areas of our business could it help?”
There are many examples of media groups effectively experimenting systematically, from Swedish radio stations using AI to develop an algorithm for running orders to Il Sole 24 Ore in Italy using it for news gathering and revenue collection.
“They do their research, make sure they got the skills and then make sure that they approach it in a sensible way by starting small. They’re using it to prototype it, like you would with any product. You wouldn’t just introduce a new journalist without making sure they were going do a good job. It’s the same with AI.”
Watch edited highlights of our interview with Professor Beckett below.