Large language models for reducing clinicians’ documentation burden

Evaluation of a clinical summarization method based on GPT-4 suggests that such models might reduce the documentation burden on clinicians — but prospective evaluation with high-priority tasks will be the true test of its potential.