Autonomous AI Should Replace Human Doctors, Billionaire’s Son Argues
Health researchers and entrepreneurs argue in a new research paper that the future of top-tier medical care won’t come from doctors working in tandem with artificial intelligence, but will instead be administered purely from autonomous AI models that are likely to outperform both physicians and human-AI hybrid teams as soon as 2030.
Bioethicist Ezekiel Emanuel, research fellow Abe Baker-Butler and Curai Health CEO Neal Khosla (son of billionaire venture capitalist Vinod Khosla), wrote in a JAMA-published paper Monday that they disagreed with the prevailing view AI should play only a "supportive role in clinical decision-making," as the American College of Physicians advises.
Instead, the authors cite research that found Google's Articulate Medical Intelligence Explorer was “significantly better” than human doctors at eliciting patient complaints , reviewing symptoms and taking medical history , the paper reported.
Another cited study found ChatGPT o3 correctly identified the final diagnosis first in 60% of 377 complex real-world cases, while 20 internal medicine physicians did so in only about 16% of a 302-case subset, according to the paper.
The authors argue against the idea of physicians working alongside an AI system, suggesting humans will create more problems than they’ll solve, citing a 2024 study that found introducing real doctors to AI-driven case studies significantly worsened results compared to AI acting alone.
The paper also argues AI models will only get smarter while human doctors are set to lose intelligence over time as they fall victim to "AI-induced deskilling,” or the gradual loss of human abilities and expertise due to overreliance on AI.
If fully autonomous AI doctors will ever be legal. The paper’s authors noted a handful of liability, reimbursement and regulatory hurdles remain unresolved and, as of now, won’t allow for non-human-doctors to administer care on their own.
Robert Wachter, one of American medicine's most prominent voices on AI, argued in his 2026 book "A Giant Leap" that AI-aided physicians will remain the best of the best in medicine while the “economy class” of patients will receive AI-only care. The American Medical Association and American College of Physicians both insist AI must remain in a "supportive role” to “ensure maximum benefit and minimum harm to patients.”
The authors compared AI’s place in medicine to its introduction to the game of chess. Deep Blue, a custom supercomputer built by IBM for chess-playing, beat world champion Garry Kasparov in a game in 1997. For more than a decade after, human-AI teams ruled the sport until AI alone blew past those hybrids by 2017. The authors think medicine is on the same curve, with one twist chess grandmasters didn’t face: the longer doctors lean on AI for cognitive work, the more their own clinical skills erode—making the eventual handoff harder to reverse.
Neal Khosla is CEO of Curai Health, an AI-driven primary care company. His father, Vinod Khosla, is a venture capital billionaire who invests in experimental technologies like biomedicine and robotics. Vinod Khosla, now worth an estimated $13.5 billion, was one of the first venture capitalists to invest in OpenAI, an affiliation that was noted among the new paper’s conflict of interest disclosures. Vinod Khosla recently led an investor group that bought the Super Bowl champion Seattle Seahawks at a $9.6 billion valuation.