Google is developing a new generative artificial intelligence (AI) tool called Genesis, which has been reportedly tested for writing news articles. According to The Guardian, the AI technology synthesizes information from various events and sources to produce news stories.
In July of 2023, Google was in talks with major American news organizations. A report by The New York Times said Google met with Wall Street Journal, Washington Post, and New York Times executives to pitch and discuss the technology.
Jenn Crider, a Google spokesperson, provided comments to The Verge, pertaining to the recent situation: “In partnership with news publishers, especially smaller publishers, we’re in the earliest stages of exploring ideas to potentially provide AI-enabled tools to help journalists with their work.”
Google maintains that Genesis is not intended to replace journalists. The spokesperson said the AI aims to assist journalists by generating headlines and different writing styles to increase user productivity.
“Quite simply these tools are not intended to, and cannot, replace the essential role journalists have in reporting, creating, and fact-checking their articles,” Crider added.
The potential abilities and limitations of Google’s Genesis are not currently available to the public. However, as anonymously disclosed to The New York Times, some of the executives in the meeting felt the pitch was “unsettling.” Two people noted that Google’s pitch took for granted the AI’s effort that “went into producing accurate and artful news stories.”
This news comes after an agreement between Associated Press (AP) and Open AI, the maker of Chat GPT: Open AI has access to AP’s archive of news stories for the purpose of training AI models. Large Language Models (LLMs) such as Chat GPT require vast amounts of written data to train the models. This allows the AI to perform various natural language processing tasks such as text generation, text categorization, conversational questions, and translation between languages. The financial terms of this deal are not publicly disclosed at this time.
Angela Misri is a journalist, novelist, digital director at The Walrus, and assistant professor at Toronto Metropolitan University. She predicts that new AI technologies like Genesis could aid journalists with some daily “grunt work” activities, including idea generation, headline writing, transcription, and basic copy editing.
Boris Babic, an assistant professor of philosophy and statistics at the University of Toronto who specializes in machine learning and AI, says it’s difficult to determine if AI tools will become a regular part of a journalist’s job. Despite his apprehension, he thinks it’s very possible.
“The time constraints on journalists and the ease of use of applications like Genesis suggests (unfortunately, from my perspective) that applying them may well become the path of least resistance in the next 5 years,” says Babic.
Babic thinks journalists should actively resist AI tools because it undermines their fundamental “integrity and independence.”
“The best tasks to automate are tasks that can be thought of in terms of solving optimization problems. In general, journalism is not one of them. The values and judgments that go into responsible reporting are not so easily substituted,” he adds.
Similar to Babic’s comments, Misri says AI technologies pose ethical challenges for journalists, arguing that newsrooms need to be transparent when using AI; she thinks new Journalistic Standards and Policies (JSPs) around AI use would create greater audience trust. Misri cites a Maru Group survey to illustrate Canadians’ interest in AI disclosure and transparency in the newsroom.
“A recent survey conducted by Maru Public Opinion for the Canadian Journalism Foundation revealed that 92 per cent of the 1,516 Canadians randomly polled “believe that news organizations should have a clear and transparent policy on how they use artificial intelligence ‘AI’ technology to produce news and information.”
The survey also found that 85 per cent of Canadians are concerned that AI use in journalism will produce or spread misinformation, whereas 86 per cent of Canadians feel that such AI use may lead to inaccuracy.
While speculative news around Google’s Genesis has remained stagnant for now, Misri hopes — with evidence like the Maru survey — that newsrooms and media organizations can “get ahead of this and put it into our JSPs as soon as possible.”