Predictive AI Expected to Surpass Physicians in Core Clinical Functions by 2030, Son of Venture Capitalist Claims

Topline

Health researchers and entrepreneurs contend that future high‑quality medical care will be delivered exclusively by autonomous AI models, which are expected to surpass both individual physicians and human‑AI collaborative teams by 2030.

Key Facts

Bioethicist Ezekiel Emanuel, research fellow Abe Baker-Butler, and Curai Health CEO Neal Khosla—son of venture capitalist Vinod Khosla—published a JAMA paper on Monday stating that they reject the prevailing notion that AI should be limited to a supportive role in clinical decision‑making, as recommended by the American College of Physicians.

They cite research showing that Google’s Articulate Medical Intelligence Explorer outperformed human physicians in eliciting patient complaints, assessing symptoms, and obtaining medical histories, according to the paper.

Another study reported that ChatGPT‑4 correctly identified the initial diagnosis in 60% of 377 complex real‑world cases, whereas 20 internal medicine physicians achieved this only about 16% of the time in a 302‑case subset, as noted in the paper.

The authors contend that physicians collaborating with AI systems may introduce more problems than they resolve, referencing a 2024 study in which the addition of human doctors to AI‑driven case studies markedly degraded outcomes compared with AI operating independently.

The paper also argues that AI models will continue to improve, whereas human doctors may experience declining expertise over time—a phenomenon known as AI‑induced deskilling, caused by overreliance on automated tools.

WHAT WE DON’T KNOW

Liability, reimbursement and regulatory hurdles remain unresolved, and currently preclude non‑human doctors from delivering care independently. The authors raise the question of whether fully autonomous AI physicians will ever be legally permitted.

CONTRA

Robert Wachter, a leading voice on AI in medicine, argued in his 2026 book *A Giant Leap* that AI‑augmented physicians will continue to represent the highest standard of care, while patients classified as “economy class” may receive care delivered solely by AI. The American Medical Association and the American College of Physicians maintain that AI should remain in a supportive capacity to maximize patient benefit and minimize harm.

SURPRISING FACT

The authors liken medicine’s trajectory to that of chess, noting that IBM’s Deep Blue defeated world champion Garry Kasparov in 1997. For many years thereafter, human‑AI teams dominated the sport until AI alone surpassed those hybrids in 2017. They suggest that medicine follows a similar path, with the added challenge that physicians’ prolonged reliance on AI can erode their clinical skills, complicating a reversal of decision‑making authority.

Source link