Why AI’s tendency to agree can be bad for learning

This season Libby Hills from the Jacobs Foundation and AI researcher Owen Henkel continue to speak with leading researchers, practitioners and educators on the Ed-Technical podcast series about the cutting edge of AI in education. They will break down complex AI concepts into non-technical insights to better understand what the research says and help educators sift the useful insights from the AI hype. 

In this short, Libby and Owen discuss recent research from Anthropic looking at sycophancy – the tendency to agree with users – in large language models (LLMs), and key research from educational psychology about how important feedback is for learning. Libby and Owen connect the two papers and explore why sycophancy is especially a problem when it comes to using LLMs for educational purposes. 

Links

Anthropic paper on sycophancy in language models 
John Hattie and Helen Timberley’s paper, The Power of Feedback