Health & Science Briefing: November 2026, When AI Medicine Met Its Evidence Problem

November was the month the health AI conversation grew up a little. For years the pitch was simple: these tools are coming, they will save time, they will catch what humans miss. This month brought a harder question into focus. Now that AI is embedded in oncology workflows, pediatric documentation, cognitive screening, and clinician search tools like OpenEvidence, who is actually checking whether it works, and by what standard? A JAMA study on OpenEvidence and a Lancet paper scrutinizing Google's AMIE chatbot both landed with the same uncomfortable subtext: peer review alone does not settle whether an AI tool is safe for a patient exam room. Johns Hopkins' new partnership with OpenEvidence suggests health systems are not waiting for perfect answers before building anyway. Beneath that AI story sat a quieter, more structural one. Rural hospitals in Minnesota described themselves as running out of road, with more than a quarter losing money on care delivery even before federal funding cuts land. Primary care's shortage numbers kept climbing, with tens of thousands of missing physicians projected by 2036. Artiva Biotherapeutics halted its lead cell therapy program after two patient deaths, laying off nearly its entire staff, a stark reminder that the cell and gene therapy field's scientific promise still runs ahead of its financial and safety guardrails. Science stories outside the clinic offered some relief from that tension. Frozen pig kidneys came back to life after months in storage, hinting at a real fix for the organ shortage's geography problem. Planarian worms refused to cooperate with a sixty-year-old memory transfer experiment, seabirds turned out to get jet-lagged just like frequent flyers, and physicists kept revisiting how chirped pulse amplification and a stubborn radio hiss once rewrote entire fields. These stories were a reminder that basic curiosity-driven research still produces some of the most durable insights in science. Taken together, the month described a system trying to modernize on two tracks at once: deploying AI and advanced therapies faster than ever, while its financial and evidentiary foundations, rural hospitals, primary care capacity, validation standards, visibly strain under the weight of that ambition.