AI models are using material from retracted scientific papers
13 hours ago
- #retracted papers
- #scientific research
- #AI reliability
- AI chatbots sometimes use retracted scientific papers to answer questions, raising concerns about their reliability.
- Recent studies show that AI tools like ChatGPT may not recognize retracted papers, leading to potentially misleading answers.
- Public reliance on AI for medical advice and scientific summaries increases the risk of spreading flawed research.
- Some AI research tools, such as Elicit and Consensus, have started integrating retraction data to improve accuracy.
- Retraction databases are not comprehensive, and publishers use varied labels for retracted papers, complicating detection.
- Experts suggest including more context, like peer reviews and critiques, in AI responses to improve reliability.
- Users and creators of AI tools must remain skeptical and perform due diligence to ensure accurate information.