Great debate: artificial intelligence will replace much of what cardiologists do

Strengths of AlGenerative language models can provide personalized instruction Al can teach procedural skills Automated image acquisition, computation, and interpretation can reduce workload Al models synthesize a wealth of medical literature AI-based decision support can guide physicians through diagnosis and management Synthetic data can be developed for predictive modelling Forms and notes can be generated to reduce administrative burden Generative Al can communicate with patients in several languages and demonstrate empathy Limitations of Al Al cannot e ectively model bedside manner or empathy in real time Oversight of medical education by physicians is essential Physicians are liable for Al-generated reports Oversight of model design and training data needed Al decisions use opaque logic Physicians must assess for bias/falsi cations in decision-making Synthetic data cannot replace randomized trial data Reports must be reviewed for accuracy, interpretability The patient-physician relationship cannot be replicated Large language models can exhibit bias Medical education Interpretation of diagnostic tests Clinical decision-making Data, image, and report generation Patient communicationThe role of AI in cardiovascular care.Artificial intelligence offers powerful tools to assist in the care of cardiac patients, but it remains to be seen if cardiologists will be replaced.Artificial intelligence can assist in medical education, automate image interpretation, and synthesize large quantities of data from a variety of sources.However, physician oversight is needed to ensure that reports generated are accurate, that bias is minimized, and the care provided is safe and inclusive.