AI could surpass doctors in medical diagnosis by 2030

Friday 21 - 22:00

By: Azzat Manal

AI could surpass doctors in medical diagnosis by 2030

Artificial intelligence is moving toward a potentially transformative role in healthcare, with recent research suggesting that increasingly autonomous AI systems could outperform doctors in certain complex diagnostic tasks by 2030.

A paper published in the Journal of the American Medical Association (JAMA) by Ezekiel Emanuel, Abi Beker-Butler and Nikhil Khosla argues that advances in AI could eventually challenge the traditional view that such systems should remain strictly supportive tools for clinical decision-making.

The researchers point to growing evidence of improvements in the ability of AI models to analyse symptoms, process medical histories and identify possible diagnoses. One example is AMIE, an experimental system developed by Google, which has demonstrated strong performance in tasks such as gathering patient information, analysing symptoms and conducting elements of medical history-taking.

Other research has also highlighted the potential of advanced AI models in complex medical reasoning. In one study cited in the discussion, OpenAI's o3 model reportedly reached the correct final diagnosis in around 60% of 377 challenging medical cases. By comparison, a group of 20 internal medicine specialists achieved a success rate of approximately 16% in a subset of 302 cases.

These findings could encourage a reassessment of how AI is integrated into clinical practice. Advanced systems can process large volumes of medical information rapidly and connect symptoms, patient histories and possible diagnoses in ways that may be difficult to reproduce within the time constraints of routine medical care.

The researchers also question whether combining doctors and AI will always produce better results than allowing AI systems to operate independently for particular tasks. Previous studies have suggested that human intervention does not necessarily improve AI-supported performance in every situation and, in some circumstances, may even reduce overall accuracy.

At the same time, greater reliance on automated systems could create new challenges for medical professionals. If AI increasingly performs diagnostic analysis and other cognitive tasks, doctors could have fewer opportunities to practise certain skills acquired through years of education and clinical experience.

This raises an important question about the future of healthcare: will AI remain primarily a tool that assists physicians, or could some medical decisions gradually be delegated to autonomous systems?

A shift toward greater AI autonomy would not necessarily mean eliminating doctors. Human professionals would still play an important role in communicating with patients, considering ethical questions, managing uncertainty and understanding social and emotional circumstances that cannot be fully captured by medical data.

The development of increasingly capable AI systems could therefore lead less to the disappearance of doctors than to a fundamental transformation of their responsibilities. By 2030, the balance between human expertise and machine intelligence could become one of the central questions shaping the future of medical diagnosis.