PhD Student Explains Study Marking Turning Point in AI and Medicine

PhD student Thomas Buckley co-authored a study that suggests AI is good enough at diagnosing complex medical cases to warrant clinical testing.

The study, led by physicians and computer scientists at Harvard Medical School and Beth Israel Deaconess Medical Center, found that a large language model outperformed physicians across many clinical reasoning tasks. The results make the case that medical AI is ready to be studied the same way as all new medical interventions: through carefully controlled, rigorous, prospective clinical trials in real care settings.

While studying machine learning and computer science as an undergrad at UMass Amherst, Buckley was drawn to how large language models could tackle problems that seemed impossible. Two years ago, he joined the first cohort of students in the AI in Medicine PhD Track at HMS.

“I love working in a field that’s going to have, potentially, a real impact using machine learning,” Buckley said.

Видео PhD Student Explains Study Marking Turning Point in AI and Medicine канала Harvard Medical School