LONDON: Google has said the company was actively investigating the application of artificial intelligence tools for news article writing.
They are also engaging in discussions with various news organisations to explore the possibility of using these tools to support journalists in their work, according to Reuters.
Although the Google spokesperson did not disclose the names of the publishers involved, sources from the New York Times claim that Google has been in talks with renowned media outlets such as the Washington Post, News Corp (owner of the Wall Street Journal), and even the New York Times itself, alongside others.
These AI tools could help journalists with options for headlines or varying writing styles, for instance, in a way that “enhances their work and productivity,” the Google spokesperson said, adding it was in the “earliest stages of exploring ideas”.
“Quite simply these tools are not intended to, and cannot, replace the essential role journalists have in reporting, creating, and fact-checking their articles,” the spokesperson said.
However, some executives who saw Google’s pitch described it as unsettling, the NYT said, adding the executives asked not to be identified. The AI tool that was pitched is called Genesis internally at Google, the NYT said, citing people familiar with the matter.
A News Corp spokesperson declined to comment on the NYT report or the AI tool, but said, “We have an excellent relationship with Google, and we appreciate (Google CEO) Sundar Pichai’s long-term commitment to journalism.”
The NYT and Washington Post did not immediately respond to Reuters’ requests for comment outside regular working hours.
The news comes days after the Associated Press said it would partner with ChatGPT-owner OpenAI to explore the use of generative AI in news, a deal that could set the precedent for similar partnerships between the industries.
Some outlets are already using generative AI for their content, but news publications have been slow to adopt the tech over concerns about its tendency to generate factually incorrect information, as well as challenges in differentiating between content produced by humans and computer programmes.
JournalismAI, a project of the London School of Economics run by Professor Charlie Beckett, supports newsrooms in their efforts to responsibly utilise AI, according to a Sky News report published earlier this month.
Since it started in 2019, but now that editors and reporters are beginning to understand the power of generative AI, its purpose has never been more obvious.
He says “a lot of newsrooms are thinking through what they might do” with the technology, but all are conscious of the pitfalls. It is evident that artificial intelligence is not yet ready to take the place of actual journalists after CNET discovered errors in an AI-written story and a hoax column in The Irish Times.
“If you get a tiny thing wrong at Sky News, people are laughing at you, it’s all across social media, and the chances of AI doing that are very high,” says Prof Beckett. “If we all get lazy and expect GPT to write our stories and scripts and so on, they may get worse.”
However, AI appears destined to have a similarly significant impact, much like how smartphones and Google search revolutionised journalistic work.
Prof Beckett predicts a smaller newsroom with AI replacing interviewing, scripting, and online stories.
This will create hybrid jobs, requiring hybrid tech and editorial skills. The savings could be used to improve human journalism, allowing reporters to interview more people and produce more imaginative, empathetic, and opinionated stories.
The future of newsrooms
The University of Kent’s Centre for Journalism examines preparing future reporters for AI-powered newsrooms.
Professor Ian Reeves says while there are “reasonable and ethical uses for it within a newsroom”, AI is also “perfectly capable of spitting out utter nonsense with a completely straight face”.
“We’ve noticed in some of the journalism assignments that we’re giving to our students that they’ve attempted to use this tech to deliver journalistic content – in some cases with rather hilarious results,” he says.
“In one piece about The Sun newspaper and its coverage of an event, the chatbot hadn’t been able to distinguish between the newspaper and the fiery star in the sky.
“[So] we’re also trying to demonstrate to them that the risks of relying on it to produce sensible content are pretty high.”
Where does AI prove useful in journalism?
Prof Reeves also believes generative AI’s usefulness lies in the fundamental journalistic tasks of talking to real people, bearing witness, and holding power to account. Google is often the first port of call for researching unfamiliar topics, but AI cannot do these fundamental tasks.
These skills will become even more important for journalists to survive in the era of AI.
“It comes down to trust and credibility,” he says. “The best journalists, the ones who make a difference, are the ones out there talking to people about how things are affecting their real lives, bearing witness to events. The ones who have the skills to find things out and reveal stuff that powerful people don’t want to be revealed.”
“That’s not something that AI can do,” he added.
He also believes that AI will replace journalism jobs that lack these skills and are considered content farm jobs, resembling journalism without interaction. Publishers may benefit from AI platforms rather than hiring people for these jobs, according to him.
Prof Reeves also shared that young people demand personalised newsroom content tailored to their format, tone, style, and platform. Reformatting and customising content, translating into different languages, and providing simple or explainer versions are crucial areas for success.
“My sort of sci-fi vision for this is a kind of Robocop journalist with all these tools to help them be more efficient and much more powerful and able to research much better, and then a content creation thing that takes their original piece and turns it into all sorts of iterations.” he said.
“And then your audience sitting at home having breakfast watching Sky News, they can get into the car, and it continues as audio with a selection of stories they’re interested in,” he added.
“Then they get home from work in the evening and just want a nice long read, and it all happens semi-automatically, where people have a kind of Spotify-like ability to shape what they get,” he continued.