Video describes a 2026 paper on bias in AI biological age clocks

Can an AI really tell how old your body is? 🧬 In this video, I break down a new 2026 research paper from University College London and Moorfields Eye Hospital on "biological age clocks", AI models that estimate your age from eye photos, brain MRI, chest X-rays, abdominal CT scans, and blood tests. The gap between your AI-estimated age and your real age is often used as a sign of how fast you're ageing. But this study found a hidden prediction bias called differential regression to the mean, which can weaken or even flip the link between the age gap and disease in older people. In this video you'll learn: ✅ What a biological age clock and "age gap" are ✅ How AI estimates age from eye, brain, chest, abdomen, and blood data ✅ What regression to the mean is, in simple words ✅ Why the bias is stronger in unhealthy people ✅ Why the age gap can stop predicting disease in people over 70-80 ✅ Why calibration and balanced training data don't fix it ✅ Why these markers work for populations but not for individuals ✅ What researchers recommend doing instead 📄 Source paper: "Prediction bias in biological ageing markers" (Lin et al., 2026), arXiv:2609.30322 (please check this ID against the PDF's first page before publishing) If you enjoyed this video, please like 👍, subscribe 🔔, and comment: would you want to know your biological age? #BiologicalAge #AIinHealthcare #Ageing #MachineLearning #Longevity Disclaimer: This video is an educational explainer of a research paper and is not medical advice. Please refer to the original paper for full details and talk to a doctor about your health.