AI chatbots offer problematic medical advice half the time, say scientists |
We are increasingly turning to AI chatbots for help. From basic research to emotional support, people are talking to AI friends in a lot of ways. But as per a damning research, you don’t want to trust AI with medical tips. Scientists tested Google's Gemini, DeepSeek, Meta AI, OpenAI’s ChatGPT, and xAI’s Grok to test their accuracy. They discovered that half (50%) of the responses were problematic. “They do not reason or weigh evidence, nor are they able to make ethical or value-based judgments,” the researchers said about AI chatbots and their medical skills. |
No comments:
Post a Comment