“News farms” in the age of artificial intelligence
As the use of Artificial Intelligence (AI) increases in intensity and diversity, so do the unknowns to be resolved by journalism in the complex coexistence with the most recent technological innovation in the digital world. What was initially seen as a disturbing possibility has now become a reality, after the News Guard organization revealed the existence of 49 online pages (the so-called “content farms”) specialized in the distribution of texts tagged as news produced by artificial intelligence software. .
These 49 content farms deliver a daily avalanche of approximately 1,200 texts, almost half of them distributed by Google Ad to an audience roughly evaluated in several tens of millions of people around the world. The target audience for these texts is, preferably, blog owners, influencers, online ideological activists ( mostly extreme right groups) and editors of journalistic pages on the internet.
The “content farms” are part of the 349 sites identified, so far by NewsGuard, that produce and distribute automatically generated information in at least seven different languages, including Portuguese. The organization specializing in website monitoring, such as the Center for Media, Data and Society, claims that “farms” are today the primary source of the distribution of fake news and misinformation.
The race for databases
That alone is enough to put the exercise of journalism in check in this new informational context that has emerged in the digital space. But there are other initiatives that are no less disturbing. Google, OpenAI, Microsoft and Apple are in an advanced stage of developing artificial intelligence software aimed specifically at news production. Google is already working with The New York Times.
OpenAI has entered into an agreement with the Associated Press to use the agency’s news archive to automatically produce news stories. The company that created the famous ChatbotGPT software also created partnerships with the American Journalism Project to develop programs for the automatic production of local news. For its part, Microsoft, which is one of the sponsors of Open AI, develops specialized software for unmanned coverage of meetings, lectures, congresses and seminars. And Apple is mysterious about its artificial intelligence projects, but it is known that it works with the financial website Bloomberg.
These are just the most popular projects of the great technological conglomerates involving the application of artificial intelligence in journalism. According to the consulting firm KPMG, no less than 43% of the functions currently performed by journalists should be automated thanks to algorithmic procedures fed by mega-banks of digital data.
AI as a challenge for journalism
This reality places journalists in front of a series of immediate challenges, created by the frantic advance of technological innovations and the impetus of investors interested in maximum profitability. The first thing my fellow journalists need to understand is that artificial intelligence is a process that is a few years old, not a new product. It seems like a subtle difference, but it’s not.
The automatic production of texts began at least 30 years ago and its growth is associated with the increase in the volume of digitized data and the permanent advance in the sophistication of the algorithms that process these same data and transform them into information. Thus, is easy to deduce that the dystopian scenarios created by marketers obey the strategies of quick commercialization of artificial intelligence products developed by the so-called Big Techs, as is the case of ChatbotGPT.
Journalism is not at the door of a catastrophe as many have claimed, but it is undeniable that our activity needs, once again, to review its procedures, rules and values in order to adapt to new professional situations. The first and most controversial of the changes is the one that will charge journalism with greater emphasis on the analysis, evaluation and recommendation of news than on the mining of raw news material. The quality of a professional’s intellectual performance will prevail over his ability to perform routine activities.
The universe of more than 5.5 billion internet users around the world creates the possibility that any fact or event will quickly end up being integrated into the flow of news on internet social networks. It is no longer worth sending a reporter to cover a coup d’état in Africa or Latin America because the news will soon be in more than one newspaper, blog or news site on the Web. The problem will be how to distinguish reliable, timely and relevant news from a flood of AI-processed material.
News curation
Working as a news curator must become a more journalistically valued activity than “shoe leather reporting” for chasing newsworthy data, facts and events. This will place on professionals a demand for intellectual preparation much greater than that existing in the analogue era. This will inevitably end up causing changes in training courses for journalists, which are still largely outdated in relation to technological advances in the field of information and communication.
The automation of the news production process offers advantages in the diversification, speed and updating of news, but does not guarantee reliability and credibility because artificial intelligence works from algorithms, which are programmed micro robots. Everything they produce depends on what they’ve been given. Wrong data generate wrong information, so inevitably someone will have to carry out a minimal check of what comes out of a “content farm”, as the organization NewsGuard does.
But the volume of material distributed is such that not even a hundred similar entities will be able to account for everything that circulates on the internet, which by the end of 2023 should house 120 trillion gigabytes of data, according to the Statista website, with a yearly average growth of 26%. These data are the so-called “pasture” where the artificial intelligence algorithms are fed.
Journalism will never be able to compete with these algorithms in terms of speed, diversity and amount of data collected online. This is what frightens many companies and institutions that fear the challenges and uncertainties of technological innovation. But journalists who bet on their professional qualifications have all the conditions to face artificial intelligence, not as a threat, but as an incentive for a more creative production of news.
