We appreciate the interest in our original article, “Performance of the predictive criteria of the American Society for Gastrointestinal Endoscopy in the diagnosis of choledocholithiasis at a secondary care public hospital in the state of Nuevo León, Mexico”,1 shown by Castrillón-Lozano et al. Our article evaluated the performance of the predictive criteria proposed by the American Society for Gastrointestinal Endoscopy (ASGE2 in 2019 for predicting choledocholithiasis at a secondary care public hospital that does not have access to magnetic resonance cholangiography or endoscopic ultrasound.

Recent advances in artificial intelligence (AI) have given rise to generative models capable of providing accurate and detailed text-based responses to written prompts (“chats”). These models have obtained high scores on standardized medical exams.3

Generative AI is a promising complement to human cognition in the diagnostic process. Nevertheless, agencies, including the US Food and Drug Administration (FDA), have currently issued important warnings, with respect to these modern differential diagnosis-generating models.4 Research, such as that by Castrillón-Lozano et al., aids in investigating possible biases and diagnostic blind spots in the generative models of AI.