Grok chatbot icon. Photo: REUTERS/TTXVN
Reuters Institute director Mitali Mukherjee said the research center's annual survey found for the first time that a large number of people use chatbots to search for headlines of articles and get news updates.
Specifically, the Reuters Institute report stated that in a survey conducted by YouGov on 97,000 people in 48 countries, only 7% used AI to search for news. However, this rate is higher among younger groups, with the figures being 12% of people under 35 and 15% of people under 25, respectively. ChatGPT by OpenAI (USA) is the most widely used chatbot, followed by Gemini by Google and Llama by Meta.
Survey respondents appreciated the personalized, tailored news delivered by chatbots.
According to the report, 27% of respondents said they use AI to summarize news; 24% use it for translation; 21% look for article recommendations, while nearly 20% ask questions about current events. The survey also noted users' skepticism about AI, with many expressing concerns that the technology could make news less transparent, less accurate, and less trustworthy. The Reuters Institute report also pointed out that traditional media - including TV, radio, newspapers, and news sites - are losing market share to social networks and video sharing platforms.
Nearly 50% of survey respondents aged 18-24 said social media like TikTok is their main channel for accessing information, especially in developing countries like India, Brazil, Indonesia and Thailand.
The Reuters Institute added that many people are still using Elon Musk’s social media platform X to get their news, with about 23% of people using it in the United States, up 8% from the 2024 survey. The increase was also seen in countries like Australia and Poland. In contrast, rival social networks like Threads, Bluesky and Mastodon have had little impact globally, with less than 2% of users using it to get news.
The Reuters Institute's annual report is considered important to those who follow the development of the media.
Instead of being programmed, today’s powerful AI large language models (LLMs) are trained on large amounts of data from the web and other sources — including news media like text articles or videos. Once trained, they can generate text and images to answer users’ natural language queries. However, these AI models are prone to problems such as “illusionism” — a term that means the AI creates information that matches patterns in its training data but is not true. Another problem with AI is that many media outlets have filed copyright infringement lawsuits against AI makers, accusing them of illegally using their content — for example, the New York Times sued OpenAI, the developer of the popular ChatGPT chatbot.
According to MINH TAM (VNA)/Tin Tuc and Dan Toc Newspaper
Source: https://baovanhoa.vn/nhip-song-so/tri-tue-nhan-tao-gia-tang-xu-huong-dung-chatbot-ai-de-cap-nhat-tin-tuc-143821.html
Comment (0)