On November 25, the Press Department (Ministry of Culture, Sports and Tourism) organized a training course on effective digital transformation models of the press.
Sharing at the program, Mr. Nguyen Hoang Nhat - Deputy Head of the Nhan Dan Newspaper Electronics Department shared the topic "Artificial Intelligence (AI) for Journalists" with trends, tools and methods of applying AI that are changing news production activities in news agencies today.
AI - tool ecosystem supports journalism processes
During the training session, Mr. Nhat introduced the AI tool ecosystem that is effectively supporting reporters and editors, from ChatGPT, Midjourney to Gemini..., helping to search, synthesize data and produce many different content formats.
AI not only saves time but also opens up new creative space when reporters can focus more on developing perspective and exploiting exclusive topics.
However, the speaker emphasized that AI needs to be seen as an assistant. This technology is very good for repetitive and time-consuming tasks, from suggesting titles, topics to writing columns based on raw data.
AI also has the ability to quickly create illustrative photos, summarize long article content, check spelling and literature, translate into multiple languages, track hot news, as well as support initial information verification.
In addition, AI stands out in its ability to analyze big data, helping news agencies evaluate reader trends and propose appropriate content strategies. AI video technology is also developing strongly, allowing the automatic creation of illustrative clips.
Not only stopping at tools, Google AI Studio now also allows editors to build AI applications themselves to suit their own processes. The AI Agent trend, where each reporter can design an automated process for daily work, is expected to soon become a standard in modern publishing activities.
prompt writing skills for journalism
If a few years ago, mastering recording tools, video recording or data analysis was considered an important skill for modern journalists, then up to now, Mr. Nguyen Hoang Nhat commented that prompt writing is "a new skill that any reporter must master".
Mr. Nhat explained that AI only truly promotes its power when reporters know how to guide it with clear, informed and structured commands. On the contrary, if prompt is vague and has general requirements, the results will be disjointed and of little value to the actual work of the publishing house.
A standard prompt from Mr. Nhat's experience includes five elements: "Scenic book" to help determine the role of the AI model, such as asking AI to play the role of a veteran journalist to provide appropriate handling. The "goal" clearly states the results to be achieved, such as building outlines, analyzing data or creating an emphasis.
Specific tasks describes the step-by-step research, reference sources or information processing that AI needs to do. "Request results" provides a specific format of the output. Finally, the attention section reminds AI to comply with important principles such as accuracy or information verification.
According to Mr. Nhat, using the # markdown structure not only makes commands clear and easy to read, but also interacts with the way many models such as GPT-4 and GPT-5 process information. Thanks to that, reporters can better control how AI interpret tasks, limiting errors in output.
He also recommends that reporters should regularly ask AI to ask questions again or suggest editing requirements to ensure the task is understood correctly by the model before starting to create content.
In parallel with writing prompt, Mr. Nhat also described the wide support that AI is providing: from title suggestions, expanding the direction of theme exploitation, to creating quick hypertesteses from raw data. AI can analyze a series of documents in a few seconds, translate a multi-linguistic meeting in real time, self-check spelling - literature, and even scan hot news according to the topic of interest of the reporter.
