In this blog post, we will examine whether robot journalism can replace journalists and the opportunities and risks that this may bring.
The world is changing rapidly with the advancement of intelligent information technologies such as the Internet of Things, big data, and artificial intelligence. These changes have sparked the Fourth Industrial Revolution by bringing about innovations in various fields, including society, economy, and industry. The term “Fourth Industrial Revolution” is already commonly found in numerous media outlets, and there is no shortage of analyses and predictions about it. The Fourth Industrial Revolution is manifesting itself in various forms in many fields, and the media is no exception. In the media field, a technology called “robot journalism” has emerged. This is a compound word combining ‘robot’ and “journalism,” and refers to “a process in which computer algorithms are involved in all stages of news article creation, with robots (software) classifying, organizing, and interpreting information to write articles on their own.”
Robot journalism broadly refers to the involvement of computer algorithms in the entire journalism process. It replaces specific tasks in the news gathering, writing, editing, and distribution processes with algorithms based on predetermined rules. In a narrow sense, robot journalism refers to “automatically writing news articles based on predetermined algorithms when specific data is provided,” and when we talk about “robot journalism” in general, we are referring to these automatic news writing algorithms. In this blog, we will discuss robot journalism in this narrow sense.
The news creation process of robot journalism is as follows:
1. Data collection
2. Event extraction
3. Selection of important events
4. Determination of the tone of the article
5. Creation of news articles
First, in the data collection stage, data to be analyzed is gathered. Next, in the event extraction stage, the collected data is analyzed to find meaningful information. In the important event selection stage, each event is weighted and evaluated for importance. After that, the context of the event is understood, a narrative perspective is established to determine the tone of the article, and then appropriate sentences are selected and arranged in order to generate a news article.
Robot journalism has already been introduced and is being used in Korea and overseas. Overseas, the LA Times automatically writes articles when an earthquake with a magnitude of 3.0 or higher occurs through “Quakebot,” and Forbes writes articles using an AI reporter called “Quill Engage.” In addition, the Associated Press in the US produces articles using an Amazon cloud-based platform called Wordsmith. Startups such as Narrative Science, Automated Insights, Whipbit, Uchit, and Data Miner provide robot journalism programs and similar programs to media companies and other businesses. There are also articles in Korea that utilize robot journalism, such as Financial News, Herald, E Today, and Electronic Newspaper. In Korea, during the 19th presidential election, a robot reporter named “NARe” from SBS analyzed voting and vote counting data in real time and wrote various data analysis articles, including regional classifications and comparisons with previous elections. In January 2019, a specialized newspaper called Robot Journalism Newspaper was launched, with robot reporters writing articles on various topics, including traffic conditions, performance news, weather forecasts, earthquake news, lottery results, and Blue House briefings.
Currently, robot journalism is mainly used in fields where data values are clear, such as sports, weather, finance, and securities. However, as technology continues to advance, robot journalism is likely to gradually expand its scope and be utilized in various fields. In this blog, we will discuss the risks that may arise when robot journalism develops and becomes commercialized on a wider scale.
First, let’s look at the advantages of robot journalism when it becomes commercialized. The biggest advantage of robot journalism is that it can convey only facts in articles. Unlike human reporters, robot reporters do not need to choose provocative words or stir up controversy to increase page views. Therefore, they can provide articles based on facts, rather than clickbait or provocative content. In addition, articles are written using algorithms, which speeds up the delivery of information. Robots can deliver breaking news faster than human reporters, and as technology advances, they will be able to write articles in 0.1 seconds, further increasing their timeliness. This speed can also be used to provide personalized articles. There is also the advantage of increased efficiency by having robots process repetitive information such as traffic conditions and weather. For these reasons, robot journalism can be seen as a new revenue model for media companies, with additional benefits such as reduced reporter labor costs and prevention of strikes.
However, robot journalism does not only have advantages. Robot journalism writes articles based on data, so if incorrect data is collected, there is a possibility of misreporting. Robot journalism practitioners say that as long as there are no problems with the data, there will be no misreporting, but this means that misreporting can occur if there are problems with the data. In an interview with a media company, a senior reporter expressed concern, saying, “In the past, 1 billion won was mistakenly written as 1 billion dollars, causing the company’s stock price to fall from its upper limit to its lower limit. Such cases can occur due to program errors.” Of course, there are fact-checking algorithms to verify the accuracy of articles written by robot journalism, but there may be problems with the fact-checking algorithms themselves, and problems may arise when the fact-checking algorithms cannot keep up with the rapid expansion of robot journalism.
Not only incorrect content, but also the lack of sufficient information can be a problem. Take, for example, a traffic accident that occurred on May 12, 2018. A driver lost consciousness while driving and scraped the median barrier on the expressway. The driver behind him deliberately caused an accident to stop the vehicle. The driver was later called a “hero” and received a commendation from the Incheon Regional Police Agency, and Hyundai Motor Group covered the cost of repairing the vehicle. If such data had not been collected and only the fact that “an accident occurred when a vehicle blocked the vehicle in front while driving” had been reported, it would not have been a good or accurate article, even if it was an objective fact.
Problems that may arise in the data collection process may also exist in traditional journalism, but robot journalism is more reliable than traditional journalism, so providing incorrect information can cause greater damage. According to a study conducted by the Korea Press Foundation, when comparing articles written by robots and articles written by journalists, people found articles written by robots to be more reliable. This shows that the mere fact that an article was written by a robot increases its reliability.
In another study conducted by Korea Research, 51% of respondents said that articles written by artificial intelligence were more reliable, far exceeding the 12% who said that articles written by journalists were more reliable. This unconditional trust can cause great damage. This damage leads to the issue of responsibility in robot journalism.
When problems such as defamation or economic damage arise due to incorrect articles written by robot reporters, it is unclear who is responsible. Unlike human reporters, robot reporters cannot be held responsible, so it is unclear who should be held responsible, such as algorithm developers or data providers. Robots cannot feel pressure, so they cannot feel the burden of high reliability or responsibility. In addition, when responsibility is distributed, accountability can also be distributed.
In summary, robot journalism is expected to deliver objective information, but there is a risk of distorted or incorrect articles due to errors that may occur during the data collection process. In addition, the algorithms used in robot journalism are designed by humans and cannot be completely fair, so problems may arise during the article writing process, such as event extraction, selection of important events, and determining the tone of the article.
This can lead to greater damage due to the high reliability of robot journalism. Another problem with robot journalism is the possibility of replacing human labor. According to the results of an experiment conducted by Dong-A Ilbo, more than half of the respondents could not distinguish between articles written by robots and those written by human reporters. This means that robot reporters have the potential to replace human reporters.
If the algorithms used in robot journalism continue to advance, it is highly likely that some of the work currently performed by human journalists will be replaced by robots.
Robot journalism will continue to expand and is likely to be commercialized on a wider scale. Although the various risks associated with robot journalism must be considered, technological advancement cannot be stopped, so users must exercise caution. When encountering robot journalism, we must avoid unconditional trust and maintain a critical stance. Even after commercialization, robot journalism will continue to develop and evolve, and the possibility of writing incorrect articles will gradually decrease. Furthermore, although it is impossible to determine whether robots will replace journalists, the spread of robot journalism will result in some journalists’ work being replaced by robots, and this scope will continue to expand.
Rather than thinking that robots will “take away” human jobs, it is necessary to view the emergence of robots as an opportunity for journalists to focus on more in-depth work and for new jobs related to robot journalism to be created. New jobs for journalists can be created by developing article writing algorithms and entrusting some of the work of reviewing articles to existing journalists. With the advancement of AI, robots are likely to take over human jobs in various fields, so we need to focus on areas that robots cannot do or find difficult to do. Robots cannot match the computing power and speed of humans, but humans have emotions, creativity, and critical thinking skills that robots do not have. Based on these abilities, we need to discover new jobs and prepare for change.
As technology advances, the world is changing to keep pace. We must understand these technologies, recognize their advantages and risks, and seek ways for humans and robots to coexist through the appropriate use of technology.
