Systematic review finds large language models are not ready for automated feedback in medical imaging education.
FO Take · Score 20
Automated feedback in medical education using large language models is a dangerous fantasy. This review exposes the chasm between hype and reality, revealing significant hallucinations, factual errors, and ethical pitfalls. Trusting LLMs with student development in critical fields is a dereliction of duty. When will educators prioritise genuine human mentorship over algorithmic delusion?
The strongest counter
LLMs offer valuable support for overburdened educators, providing instant, basic feedback and freeing up time for complex cases. Early adoption, with careful human oversight, is crucial for refining their potential.
