Google Testing New AI tool to Write News Articles
In a groundbreaking move, Google has revealed Genesis, an AI tool capable of processing current events data and writing coherent news articles. The tool has been demonstrated only to a select group of senior staff at The New York Times, The Washington Post, and News Corp.
Google stated that it is currently in the early phases of exploring an AI tool designed to aid journalists by providing options for headlines and various writing styles. The company emphasised that this technology is not meant to replace journalists but rather to support and assist them in their work.
Genesis AI was discreetly demonstrated to senior staff at select news organisations. The tool was then revealed in an exclusive report by The New York Times, itself working with The Washington Post and The Wall Street Journal in testing the new product.
With the ability to churn out news copy at a rapid pace, the tool offers productivity advantages. Google spokesperson, Jenn Crider, stated that “in partnership with news publishers, especially smaller publishers, we’re in the earliest stages of exploring ideas to potentially provide A.I.-enabled tools to help their journalists with their work.”
As Google and other tech giants delve deeper into AI’s capabilities, it becomes essential to strike a delicate balance, preserving the essence of responsible journalism while embracing the efficiencies AI offers. Striking this balance may well shape the future of the media industry — as the major players work together to harness the potential of this groundbreaking technology while safeguarding the principles that underpin responsible journalism.
At present, it is not clear what differentiates Genesis from Google’s main AI tool, Bard. Certainly, they are based on very similar, if not the same, Language Learning Model (LLM) technology. It is likely that Genesis will combine the AI functionality of Google News with an enterprise-specific version of Bard.
Such developments could lead to Google’s search algorithms favouring articles created by its AI tools. For the time being, Google has chosen not to comment on whether it would choose to implement such a strategy.
The advertising giant has until now served largely as a sort of curator — directing users to published websites for further reading. Along with making incorrect claims, AI may also fail to direct users to more credible sources. This is significant for SEO and those in the SEO industry.
A number of organisations around the world are currently exploring the possibilities of AI within their newsrooms. The Associated Press has long since used AI to write corporate earnings reports.
Taking advantage of bots to put relatively “dry” economic data into words is one thing, but using AI to produce articles across a broad gamut of news and culture topics will be unchartered territory — at least for the world’s major recognised publications.
Genesis raises ethical concerns and challenges the traditional landscape of journalism. Many in the industry are apprehensive. Anonymous sources present for Google’s pitch to senior newspaper executives described the presentation as “unsettling.”
One of the primary concerns surrounding Genesis is the potential for misinformation. AI-generated content mimics human writing styles with remarkable accuracy, but distinguishing between authentic journalism and artificially created content has become increasingly challenging.
Earlier in 2023, the American media website CNET started publishing articles using generative AI — a move which famously backfired, with the company having to amend over 50% of the content due to factual inaccuracies. The AI also had a tendency to churn out fully or partly plagiarised content.
There is a real risk of malicious actors exploiting the technology to manipulate public opinion — spreading disinformation or fake news.
Bill Gates addressed the issue in a recent blog post. Regarding AI content influencing elections, in particular, he stated: “The idea that technology can be used to spread lies and untruths is not new. People have been doing it with books and leaflets for centuries. AI takes this problem of fake text and extends it.”
What’s certain is that tech giants, news organisations, and regulators will be forced to collaborate. The EU, US, and other nations around the world have been scrambling into action to introduce policy for AI-generated content.
In democracies, transparency is going to be crucial in maintaining fair and trustworthy journalism — so as to preserve some public confidence in the media (such as it is). Dictatorships and authoritarian governments will surely find other uses.
Google must also address issues of data privacy and security as the Genesis AI collects and processes vast amounts of information. Ensuring that user data is safeguarded and not exploited for commercial gains is paramount to maintaining public trust in the technology.
Google’s new tool comes at a time when OpenAI, the parent company of ChatGPT, is establishing its own news and tech-sharing collaborations with media organisations. The Associated Press (AP) reached a two-year agreement with OpenAI a couple of weeks ago. As part of this deal, the AP licensed its text archive of news stories to help train OpenAI’s AI algorithms. In return, the Associated Press will be given access to OpenAI’s tech and product expertise.
Proponents of the Genesis AI will highlight the potential to unlock new avenues for creative storytelling and audience engagement. By automating routine tasks, journalists can focus on in-depth reporting and analysis, elevating the overall quality of news content. The creative potential of AI may unlock new avenues for storytelling and audience engagement, captivating readers in innovative ways.
Others would point to job losses for journalists, editors, and other media professionals. News organisations will need to carefully navigate the transition to AI-assisted newsrooms while ensuring that human expertise and values remain integral to the process.