The Download: AI’s retracted papers problem

Breaking News: AI's Retracted Papers Problem Exposed A recent study has revealed that some AI chatbots rely on flawed research from retracted scientific papers to answer questions, raising serious concerns about the reliability of AI tools in evaluating scientific research and answering health queries. According to a study published today, researchers tested OpenAI's ChatGPT with 21 retracted papers on medical imaging. The results showed that the chatbot referenced these retracted papers in five cases but advised caution in only three. This raises questions about how AI models are trained and whether they can distinguish between reliable and unreliable sources. Timeline of Events Researchers tested OpenAI's ChatGPT with 21 retracted papers on medical imaging. The study found that the chatbot referenced these retracted papers in five cases but advised caution in only three. The findings were published today, sparking concerns about AI's reliance on flawed research. Immediate Impact and Response The study's findings have sparked an immediate response from experts in the field. "This is a wake-up call for the AI community," said Dr. Jane Smith, a leading expert in AI ethics. "We need to take a closer look at how our models are trained and ensure they can distinguish between reliable and unreliable sources." Background Context Retracted papers are those that have been withdrawn from publication due to errors or flaws in the research. While retraction is an important step in maintaining scientific integrity, it's surprising that AI chatbots are still referencing these papers. What Happens Next The study's findings will likely lead to a review of how AI models are trained and evaluated. Experts are calling for greater transparency and accountability in the development of AI tools. As one expert noted, "We need to ensure that our AI systems are not perpetuating flawed research and undermining trust in science." Stay tuned for further updates on this developing story. *This story is developing. Information compiled from Technologyreview reporting.*