experiment · Focus on Health Professional Education A Multi-Professional Journal · la publicación, 31 mar 2026 · suscríbase
ChatGPT calificó 557 exámenes médicos. Marcó más duro que el experto humano.
Un estudio australiano comparó las notas de la máquina con las de un docente. En 63% de las respuestas, la diferencia fue demasiado grande para un examen decisivo.
Pregunte a weeklyAI
Pregúnteme por este estudio: a quiénes se estudió, qué encontró y qué no dice.
Las conversaciones se guardan mientras exista weeklyAI, para mejorar la publicación. Se responde en el idioma en que usted escribe.
En cursos de posgrado, calificar cientos de respuestas escritas consume días de trabajo docente. La idea de que ChatGPT lea y puntúe esos textos resulta atractiva. Dos autores, uno de un colegio médico de la India y otro de la Universidad de Sydney y del Hospital Westmead, en Australia, decidieron probarla.
experiment · Focus on Health Professional Education A Multi-Professional Journal · the paper, 31 Mar 2026 · subscribe
ChatGPT Marked 557 Medical Exam Answers. It Disagreed With the Human Expert on 63 Percent of Them.
In one postgraduate course, the software graded more harshly than the person who knew the subject — and the gap was widest on the questions that ask students to weigh evidence.
Ask weeklyAI
Ask me about this study: who was studied, what it found, and what it does not say.
Conversations are saved for as long as weeklyAI exists, to improve the publication. Answers come in the language you write in.
Postgraduate courses run on short written answers. A student reads a case, writes a few hundred words, and waits. Someone has to read it, mark it and hand it back. That is slow work, and it is the kind of work many teachers are now being told a chatbot could do.