ChatGPT and Google Gemini are terrible at summarizing news, according to a new study
Posted by:
John Harrison
A recent study by the BBC has revealed that popular AI chatbots, including ChatGPT, are struggling to accurately summarize news articles. The study assessed ChatGPT, Copilot, Gemini, and Perplexity on their ability to summarize 100 news stories, with 51% of responses having significant issues and 19% containing factual errors. Examples of inaccuracies included misrepresentation of facts and failure to provide essential context. The study highlighted the challenge for AI in distinguishing between opinion and fact, with Microsoft’s Copilot and Google’s Gemini showing more significant issues compared to OpenAI’s ChatGPT and Perplexity. The findings emphasize the importance of being cautious when relying on information from AI chatbots, as the technology continues to evolve rapidly. While AI progress is advancing faster than Moore’s Law, it is recommended not to solely trust AI for daily news consumption.
Recent Posts1
