Source Body Text

A calf cramp should not be a brush with death. Mine almost was. For five days, I had what felt like a stubborn muscle spasm in my left calf. It was tender, swollen and getting worse. I assumed it was a muscle problem and went to my chiropractor, who treated it as a muscle issue. But the pain kept worsening, and I turned to an AI health tool I had built for myself, based on my expertise training companies on how to adopt AI tools effectively, using my medical records, medications, lab work, and visit notes. It flagged deep vein thrombosis, or DVT, and pointed me to the diagnostic step that mattered: an ultrasound. DVT is a blood clot in a deep vein, usually in the leg. The blood clot symptoms include pain, swelling, warmth and skin color changes, especially when they affect one leg. It also said suspected DVT should be assessed urgently and, when a doctor thinks DVT is possible, the patient should be referred for an ultrasound scan quickly. So I called my primary care office, which advised me to schedule an appointment or go to urgent care. Ordinarily, that would have sounded sensible. But urgent care could not provide the scan. My doctor’s office could not provide it either. If I followed the slow path, I might lose days before being sent to the emergency room anyway. That question mattered because deep vein thrombosis can become deadly when part of a clot breaks off and travels to the lungs, causing a pulmonary embolism. The CDC describes DVT and pulmonary embolism as serious conditions that are often under-diagnosed. The National Heart, Lung, and Blood Institute warns that a pulmonary embolism can cause death when a clot is large or when there are multiple clots. So I talked to my AI tool and went to the ER, despite knowing I would spend many hours there instead of a few minutes before being seen at urgent care. The ultrasound found four clots in my left leg. Afterward, the danger stopped being theoretical. I learned that my wife’s grandfather had died from a pulmonary embolism. So had the mother of one of her close friends. What had felt like a nagging cramp suddenly looked like the edge of a cliff. This is not an argument for replacing doctors with machines. The emergency room doctors did the indispensable work. They ordered imaging, interpreted results, weighed whether to admit me, spoke with specialists and sent me home on blood thinners when it was safe to do so. The AI did not cure me. It just helped me ask the right question in time. The science is catching up with these stories. A new study on medical AI in Science, led by researchers affiliated with Harvard Medical School and Beth Israel Deaconess Medical Center, tested a large language model on clinical reasoning tasks, including real emergency department cases. Science News reported that the model was more likely than physicians to include the correct diagnosis among possible answers. That does not mean patients should trust random chatbots over clinicians. It means doctors and AI may be safer together than either is alone. A useful AI system can surface possibilities, organize medical records, help patients describe symptoms clearly and reduce the chance that a dangerous pattern gets mistaken for something ordinary. There is a real danger too. A Guardian report on AI health advice found that one in seven people in the UK are using AI chatbots for medical guidance instead of seeing a GP. That should worry everyone. A chatbot cannot examine a leg, hear breathlessness, notice distress or take responsibility for a patient. But fear should not become denial. AI in healthcare needs regulation, testing, transparency and clinical supervision. It also needs humility from institutions that too often expect patients to navigate fragmented systems alone. My lesson is not that everyone should outsource their health to software. It is that patient advocacy now has a new tool. A safe AI assistant, trained on your medical data, can help people gather their records, ask whether something urgent is being missed and push for the right diagnostic step. Medicine has always depended on second opinions. The next one may come from software, and the urgent task is to make sure it is accurate, accountable and used to save lives. Gleb Tsipursky, PhD, serves as the CEO of the future-of-work work consultancy Disaster Avoidance Experts and wrote The Psychology of Generative AI Adoption at Work (Georgetown University Press, 2026)