To survive, journalism will have to look back – 05/04/2023 – Rodrigo Tavares

I can’t remember if my typing teacher at a public school in the small town of Castelo Branco, in Portugal, in the 1990s was called Manuela or Maria. She mastered a typewriter like an instrumentalist. But she lost her job a few years later when the subject dropped out of the teaching curriculum.

When the subject of dactylographia was introduced in Portugal, in 1847, students were promised that they would be “qualified for commercial life and exercising honorable and lucrative positions on the continent, islands, Africa and Brazil”. The truth lasted about 150 years.

Are journalists the new Manuelas or Marias? Anyone who asks a deep learning algorithm like ChatGPT today to write about former President Trump’s court appearance on April 4 will be able to receive, in a few seconds, all kinds of texts (more opinionated, more analytical, factual only), with nuances and idioms in different languages. The quality, honestly, is not inferior to the articles on the same subject produced by human beings for the Brazilian press.

There are two perspectives on this phenomenon. The first rightly points out that artificial intelligence (AI) uses a pre-existing database and, therefore, is an information aggregator, not a creator. To write about Trump, ChatGPT relied (without respecting copyright) on thousands or millions of news articles published on the subject. That is, AI depends on the prior existence of journalists who create the content. It is also argued that AI could be a tool to assist the journalist, facilitating contextualization, helping him to write faster, simplifying the alignment of texts to Writing manuals, trying to find new angles to deal with a report.

We woke up to AI with ChatGPT in 2023, but already in 2018 Forbes had launched Bertie, an AI platform that learns journalists’ writing style, identifies the topics they typically write about, and provides suggestions to improve the quality of a news story (style, content, data, photos) or recommend real-time trending topics to cover.

ChatGPT agrees with this minimalist perspective on transformation. He says that “AI is currently better suited for repetitive and standardized tasks, while journalism involves human skills such as critical thinking, investigation, data analysis, interviewing, contextualization and ethical decision making”. Manuela or Maria would survive the virus of artificial intelligence.

The second prospect has already arranged his funeral.

The short-term trend is that artificial intelligence can analyze and interpret images and videos, understand the meaning of audios and write original texts. Machines don’t feel. But they will have unlimited resources to collect, evaluate and produce information, without human intervention. They will be able to interview in real time and detect inconsistencies or novelties in the information provided by the interviewee. It could also be the machines that literally show up for the content they produce.

Already in 2018, the Chinese news agency Xinhua presented its first TV anchor created by AI, a kind of algorithmic William Bonner that can command news programs for days at a time, always up to date. Human journalists, as producers and presenters of content (hard news), will be like fax machines, dial-up internet, landline telephones or typewriters.

But it will not be the end of journalism. You will be reborn in at least two dimensions.

The first is that of checking. The journalist will be the one who investigates, confronts and compares information originally produced by AI. A gatekeeper. The starting point could be the current data verification agencies, created to verify the credibility of political speeches or the veracity of information circulating on social networks. Since the emergence, in 2003, of the first platform for checking information, FactCheck.org, around a dozen similar projects have emerged in Brazil, including the Mentirômetro, of this Sheet.

In Brazil, the piauí magazine also has a checking coordination that meticulously refines all the information produced by the authors themselves, a job that goes far beyond text editing.

As early as 1913, the American Ralph Pulitzer was the first to create, in his newspaper, the extinct New York World, an internal data verification team to guarantee the reliability of the information provided to the public. The verification of information is not a minor activity of journalism, but one of its foundations.

Another dimension is that of investigative journalism, which presupposes backstage, Lacanian analysis of the meaning and signifier of each whispered word, access to reliable sources of information and cognitive interpretation of non-public information. To celebrate its centenary, Folha highlighted 100 big news stories. Almost none could have been given by an AI tool.

Newsrooms will slim down and costs will be significantly reduced, but perhaps AI will have the knock-on effect of encouraging journalism to return to its primary mission of critical truth-finding.

Today many journalists, especially the younger ones, do industrialized and poorly paid work, guided more by impact statistics than by critical thinking.

The future of journalism may imply a return to the craft of craftsmanship, that of creating objects through the transformation of raw materials using hands, intelligence and sensitivity as the main working instruments.