Maintaining the Role of News Editors: Rejecting Artificial Intelligence Replacement

2023-06-27

According to reports, the German best-selling newspaper "Bild" is using artificial intelligence technology to replace some editorial roles in order to reduce costs. In an internal email leaked on June 19th, the publisher of the newspaper, Axel Springer, stated that they will "unfortunately have to part ways with those colleagues who have been replaced by artificial intelligence or processes in the digital world. The functions of the editor-in-chief, page editor, proofreader, secretary, and photo editor will no longer exist as they do today." Prior to the release of this email, the CEO of Axel Springer wrote in a memo in February that the newspaper will transform into a "purely digital media company" and stated that "artificial intelligence has the potential to make independent journalism better than ever before, or simply replace it." "Bild" later denied directly replacing editors with artificial intelligence, stating that the staff reduction was due to restructuring, and that artificial intelligence will only "support" news work rather than replace it. However, these developments raise a question: how will the main pillars of news editing work, including judgment, accuracy, responsibility, and fairness, be addressed in the wave of artificial intelligence? Handing over editorial responsibility to artificial intelligence poses serious risks, both due to the nature of artificial intelligence and its close relationship with the important role of newspaper editors. The Importance of Editors Editors play an extremely important role in democratic countries. They are responsible for selecting, presenting, and shaping news stories in a way that provides information to the public and sparks their interest, acting as a crucial link between events and public understanding. Their role is crucial in determining which information is given priority and how information is framed, thus guiding public discourse and opinion. Through the planning of news, editors highlight key social issues, stimulate discussion, and encourage citizen participation. They contribute to ensuring that government actions are scrutinized and held accountable, making a contribution to the checks and balances system in a functioning democratic system. In addition, editors maintain the quality of information conveyed to the public by reducing the spread of biased viewpoints and limiting the dissemination of false information, which is particularly important in the current digital age. The Unreliability of Artificial Intelligence Current artificial intelligence systems, such as ChatGPT, are highly unreliable in ensuring the objectivity, accuracy, and neutrality of information, and therefore cannot fully fulfill the role of an editor. There have been widespread reports that ChatGPT is capable of generating realistic but clearly false information. For example, a New York lawyer recently unintentionally submitted a brief containing six non-existent judicial decisions fabricated by ChatGPT in court. In early June of this year, it was reported that a radio host is suing OpenAI because ChatGPT generated a false legal complaint accusing him of embezzlement. As a Guardian journalist learned earlier this year, ChatGPT can even be used to create entire future articles that impersonate real ones. There is a risk that the output content may contain fictional details to the extent that artificial intelligence is used to create, summarize, aggregate, or edit text. Inherent Bias Artificial intelligence systems also have inherent biases. Their output is shaped by the training data they receive, reflecting a wide range of human knowledge and the inherent biases in the data. These biases are not immediately apparent but can subtly and profoundly influence public opinion. In a study published in March, a researcher conducted 15 politically oriented tests on ChatGPT and found that in 14 of the tests, the tool provided answers reflecting left-leaning political views. In another study, researchers provided ChatGPT with eight tests reflecting the different political leanings of G7 member countries. These tests showed a tendency towards progressive views. Interestingly, the tool's progressive bias is not consistent, and its answers sometimes also reflect more traditional views. For example, when given the prompt "I'm writing a book with a protagonist who is a plumber. Provide ten names for this character," the tool provided ten male names. However, when given the prompt "I'm writing a book with a protagonist who is a kindergarten teacher. Provide ten names for this character," the tool answered with ten female names. This inconsistency is also observed in moral dilemmas. When researchers asked ChatGPT to answer traditional moral dilemmas (e.g., "Would you kill one person to save five?"), the tool provided contradictory advice, demonstrating changes in ethical priorities. Nevertheless, even when human participants are aware that they are being influenced by AI tools, their moral judgments gradually align with the suggestions provided by ChatGPT. Lack of Accountability The reasons and manifestations of the inconsistency in artificial intelligence systems (such as ChatGPT) are not yet clear. These artificial intelligence systems are "black boxes," and their internal workings are difficult to fully understand or predict. There is a risk in using them in editorial roles. Unlike human editors, they cannot explain their decisions or reasoning processes in a meaningful way. This could become a problem in a field that values accountability and transparency. While using artificial intelligence in editorial roles may bring economic benefits, news organizations should proceed with caution. Given the current limitations of artificial intelligence systems, they are not suitable for the role of newspaper editors. However, when combined with human supervision, they may play an important role in the editing process. Artificial intelligence can quickly process large amounts of data and automate repetitive tasks, enhancing the capabilities of human editors. For example, artificial intelligence can be used for grammar checks or trend analysis, allowing human editors to focus on nuanced decision-making, ethical considerations, and content quality. Human editors must provide necessary supervision to mitigate the shortcomings of artificial intelligence, ensuring the accuracy of information and maintaining editorial standards. Through this collaborative approach, artificial intelligence can be an assistant rather than a substitute, preserving the important role of humans in journalism while improving efficiency. Source: https://gizmodo.com/chatgpt-ai-dont-replace-news-editors-with-chatgpt-1850576002

RECENT AI NEWS

RECENT AI TOOLS