Amidst the Wave of Artificial Intelligence

After about 20 years of expanding into online media, Vietnamese press is now facing both challenges and opportunities to change as the wave of artificial intelligence explodes.

Digital transformation turning point

In the first half of the 2000s, Vietnamese press agencies took steps to expand to online formats to keep up with the times. At that time, traditional print media gradually developed electronic versions to meet the needs of readers as internet access gradually became more popular.

After newsroom convergence, news agencies are looking to adopt AI advances. Illustration: Created by AI

Along with the expansion on the internet platform, since then, the Vietnamese press has continuously improved its multimedia communication model and accelerated the digital transformation process by diversifying product forms such as: podcasts, videos, megastories, infographics, long form, data journalism, etc. Notably, the YouTube channel of Thanh Nien Newspaper has reached nearly 6 million subscribers. At the same time, the Vietnamese press has also expanded its social media channels to increase interaction with readers. Most press agencies in Vietnam now have social media accounts and the fanpages of many of them have a huge number of followers, typically the Facebook fanpage of Thanh Nien Newspaper has more than 2.2 million followers. Since then, social media has also become an effective communication channel for press agencies. Not only diversifying forms and products, the press also has to strongly transform digitally in its organization and operation, especially organizing a converged newsroom to maximize resources for all media channels: print newspapers, online newspapers, social networks, etc. Many newsrooms have applied digital technology to change their operating methods including publishing process management, data management, and public interaction management. In particular, the content management and operation system is continuously upgraded, gradually integrating new technological advances to allow the convergence of reporters and editors who can handle multiple multimedia contents at the same time. However, another reality is that many units in general, including businesses and not just press agencies, focus too much on multimedia products and content without a systematic organization based on a compatible digital platform.

Challenges from artificial intelligence

The digital transformation process is constantly evolving. In less than 2 years, artificial intelligence (AI) has exploded globally. Unlike the internet in the past, AI soon appeared in Vietnam with many remarkable advances, especially generative AI: creating content, images, videos, etc.

An AI-generated product by itself can hardly become a journalistic product without the standardization of professional journalists.

Advances in generative AI can be based on collected information and data to write content from stories, scripts, articles, etc. Some media agencies have even experimented with writing articles and creating illustrations using AI. Thanh Nien Newspaper has also experimented with illustrations and tried publishing articles created by AI. These advances have made many people ask: Can AI replace journalists? But in reality, it is not difficult to answer the question: AI cannot or has not yet replaced humans in journalism! Generative AI products are still very limited. Why answer like that? First of all, in terms of authenticity, a journalistic product must always carefully check and evaluate the authenticity of every detail of data and information. In addition, journalistic products must always comply with professional principles and legal regulations, typically fully recording information from related parties based on a network connecting information sources. Meanwhile, many of the world's leading technology experts have repeatedly voiced concerns about the authenticity of AI-generated products. Right from the source of content, data, and input information, they do not really go through all the steps of assessment. Therefore, a generative AI product itself can hardly become a journalistic product without the standardization of professional journalists. The author himself has also spent a considerable amount of time testing that. Smarter AI is still based on "machine learning" and "deep learning" that are "taught" by humans. Similarly, images or videos from generative AI can only be used for illustration to a limited extent. As for images and videos that are topical and reflect reality, the prerequisite is accuracy, so generative AI products are practically worthless in such cases. On the contrary, despite many gaps in accuracy, the explosion of generative AI products is requiring the press to face a larger and faster flow of information with a lot of fake news. That also forces the press agencies themselves to increase the control of information and data.

However, even though it cannot or has not yet been able to replace journalists, AI still has great potential to support journalists. Typically, it supports the synthesis of data, information or proposes some solutions to inspire people. This is an opportunity for press agencies to improve their work efficiency. But this opportunity requires adequate preparation in both technology infrastructure and human resources. Regarding technology infrastructure, it not only requires powerful equipment and means with greater operational efficiency, but also must ensure information security. Regarding human resources, it not only requires a team of technology-savvy human resources to develop AI applications, but also must equip journalists with knowledge so that they can effectively exploit the applications. For example, although AI is intelligent, to create a photo or perform a task to support information synthesis, users need knowledge to give appropriate commands for the application. Right from the basic generative AI applications , users who want to exploit them effectively must have knowledge of "prompt engineering" so that the AI ​​application can understand the user's wishes. To put it simply, they must learn how to "command" AI to work. All of this requires a comprehensive improvement in technology, organization and resources suitable for the AI ​​era to be able to optimize the effectiveness of this wave.

Thanh Nien Newspaper has applied AI to test some newspaper covers and illustrations.

Source : https://thanhnien.vn/giua-lan-song-tri-tue-nhan-tao-18524061416463356.htm