AI answers cancer questions not interesting for scientists
AI doesn’t always get it right when it comes to answering patients’ questions about cancer, scientists say.
In a paper published in the European Journal of Cancer, the scientists noted that there are challenges in the accuracy, reference quality, and readability of health information. These issues are particularly evident in languages other than English, where illusions remain a concern.
The study examined seven leading AI chatbots, including ChatGPT, Google’s Gemini, Microsoft’s Co-Pilot, MetaAI, Claude, Grok, and Perplexity, and assessed their ability to answer common cancer-related questions in English, Arabic, French, Chinese, Thai, Hindi, Nepali, and Vietnamese.
With millions of people turning to AI chatbots for health advice, ensuring that these tools provide accurate, understandable and well-referenced information is crucial, scientists say, and the study could be seen as a wake-up call for AI companies playing a growing and vital role in public healthcare
