New studies show that specialized AI systems can diagnose diseases and make treatment decisions as well as, or even better than, physicians in simulated patient cases. One result, however, suggests that these AI systems may not age well as their base models become outdated. Two papers published in Nature demonstrate the potential of AI in medicine, with systems like MIRA and AMIE producing accurate results. But as AI continues to advance, it’s essential to consider the long-term implications of relying on outdated technology. Why it matters: The rapid development of AI in medicine highlights the need for a more nuanced understanding of the relationship between technology and human expertise.