Artificial Intelligence, a Threat to Journalism?
Do journalist robots help journalists work? There might be some help, but robots are unable to produce in-depth coverage or detect an event that can be developed into investigative coverage.
In November 2022, a chatbot based on artificial intelligence produced by OpenAI and named ChatGPT made its debut.
This machine can answer questions in a very short time from a huge amount of data it collects. ChatGPT itself is a Generative AI-based technology product.
The conversation model created by this chat-bot makes it interesting because it answers as if two people are engaged in a Q&A. Since ChatGPT was launched, OpenAI's company value has reached US$29 billion. Several trials and comparisons between ChatGPT and Wikipedia answers showed that ChatGPT had a better performance. Technology capability is something that distinguishes the two.
Two institutions that immediately perked up their ears when they learned about the emergence of this new intelligent machine were the education and journalism institutions.
Teachers in the education sector have great concerns that this machine would facilitate the trends of cheating among students in answering various questions put to them, from drafting a paper proposal and speech text to exams.
Also read:
> Quality Journalism Needs Public Support
> Media, the Year of Politics and Games
Meanwhile, in the journalism organization, ChatGPT is also being responded to with a number of concerns. Will this machine replace the role of journalists? Will this machine be used by certain parties to enter false information into the machine? Will the role of journalists for reporting events, writing news and analyzing events be replaced by these machines?
It must be admitted that it is still too early to know, because we have not dug deeper or seen this product in special relation to the journalism institution. A number of studies should be carried out to see the direct links.
The future of trust
ChatGPT is phenomenal. Within two days after it was announced, a million people in the world downloaded it. Two months later, 100 million people downloaded the application. Every day no less than 13 million people visit the site. At the end of December 2022, the number of visitors per day reached six million people (DetikiNet).
Responding to this phenomenon, Janet Haven, executive director of Data & Society, wrote on the Nieman Lab website at the end of 2022 (Predictions for Journalism 2023), that in addition to the very impressive growth in its users, ChatGPT used for certain purposes could bring about a situation where the issue of trust was at its lowest point and could reduce existing public discussions to a homogeneous society.
Haven said it had become a necessity where the technology that was developed in society was not in accordance with what its creators imagined and the presence of ChatGPT, if it was filled with incorrect information, would produce conditions in which the people's trust in information would be damaged.
Haven also said the United States government and many other parties were working hard to regulate how misleading information must be suppressed, but ChatGPT could produce opposite conditions.
A slightly different view was shared by Eric Bergman, one of the founders of Factal. He stated that ChatGPT would not replace real journalists or the analyses produced by journalists or experts. According to Bergman, ChatGPT "hallucinates errors" and it is more appropriate to use ChatGPT as a research assistant.
Also read:New Dynamics of the Press
If ChatGPT is given input in the form of long interview results, this machine will help summarize the interview results in the form of a number of points that make the work of journalists easier.
While the work of ChatGPT-style journalists will not win any journalistic awards, Bergman also says it has two big problems: copyright and plagiarism problems, as well as problems related to propaganda and disinformation.
Van Peursen's way of thinking
Again, I use Van Peursen's way of thinking to respond to various new phenomena. The first way of looking at it is mythical: looking at something with fear, confusion and anxiety. The next point of view is ontological: how the new phenomenon is studied, how the mechanism works, what the response to it as a technology is and where the limits of its capabilities are.
The third point of view is functional: after knowing more deeply about the phenomenon or product, it is time for us to decide what we are going to do after we know the A-Zs of a technology or phenomenon.
We will also try to align our interests with this new discovery. Is this technology suitable for our lives? Is this technology relevant to our lives? Are our needs answered by this technology? To what extent can we take advantage of this technology, not to be simply amazed by the technical prowess it produces.
It is time for us to decide what we are going to do after we know the A-Zs of a technology or phenomenon.
I think Van Peursen's way of thinking is suitable for viewing various new technologies or phenomena. In Javanese culture, for example, there are the expressions ojo gumunan (do not be easily amazed) and ojokagetan (do not be easily surprised), which can actually be used here so that we are not easily amazed or afraid of new things.
On the contrary, we need to further investigate the new product or phenomenon before we ourselves will in the end determine what we want to do with the new product or phenomenon.
Do we need to worry?
A decade ago, the term “robot journalism” emerged, which was feared that it could also replace the role of journalists. Several media in Indonesia are trying to use and some news can indeed be generated by a robot: making reports on sports matches and making reports on the ups and downs of stocks on the stock exchange. More than that? Wait a minute.
Robots are unable to produce in-depth coverage or detect an event that can be developed into investigative coverage. Perhaps robots cannot write a travel report filled with nuances of literary journalism and the use of specific words to beautify the writing.
Do journalist robots help journalists work? There might be some help, such as producing live news from something that can be seen with a score or numbers.
However, a deeper analysis and linking one factor to another perhaps this still cannot be done by robots. So, from this we can say that the main factors in this technological development are still humans, who will design the direction of future development.
Also read: National Press Day, Convergence and National Economic Recovery
In data journalism, for example, the most basic program is mastering Microsoft Excel to help organize basic data, to sort it and so on. However, what is the thesis to prove from the pile of data? What is the conclusion from this search? It goes back to the journalist (humans) to determine that. Microsoft Excel is only a tool.
Today's journalism encounters many technologies, ranging from creating news based on search engine optimization (SEO), using certain algorithms to connect news and its audience and audience metrics. However, back to the controllers, where is all this progress going? In the end, the algorithm moves dynamically. It is not static.
Self-generating
Is it possible if, for example, a newspaper like Kompas collects various data that it has had for 58 years to produce a smart machine that can answer various questions about Indonesia? For example, to answer the question, "What were the events of May 1998"? Is it possible that there is a smart machine at Kompas that can provide accurate and objective answers regarding this question?
With a collection of information spanning six decades with the neat management of information and the maintenance of the information it has, Kompas should have the basic ingredients to produce a machine that can answer the questions raised. The only problem is, can Kompas have that kind of technology?
IGNATIUS HARYANTO, Lecturer in Journalism at Multimedia Nusantara University, Serpong
This article was translated by Kurniawan Siswo.