Demis Hassabis, CEO of Google DeepMind, recently asserted that artificial intelligence could replace doctors in certain roles—thanks to its speed in diagnostics, treatment planning, and data analysis. He pointed out that AI models now drastically reduce the time to develop drugs and map protein structures—DeepMind’s AlphaFold mapped over 200 million proteins in a year, a task that would have otherwise taken humans over a billion years of collective effort.

However, Hassabis emphasises that AI cannot replace nurses, whose responsibilities hinge on empathy, emotional support, and hands-on patient care. These human-centric roles require emotional intelligence something machines are not equipped to replicate.

His candid comments underscore a nuanced vision: while AI may transform physician workflows handling analysis, accelerating drug discovery, and offering decision-support nurses will remain indispensable due to their interpersonal, compassionate care and adaptability.

Indeed, a recent arXiv study comparing AI systems to board-certified clinicians in urgent care reveals that an LLM-based AI doctor matched clinical diagnostic accuracy 81% of the time and aligned treatment plans in over 99% of cases. Yet, such systems still depend on human oversight.