Victoria Song Improves Liver Markers and Cuts Triglycerides 70 mg/dL With 100-Day AI Twin
Updated
Updated · The Verge · Oct 14
Victoria Song Improves Liver Markers and Cuts Triglycerides 70 mg/dL With 100-Day AI Twin
2 articles · Updated · The Verge · Oct 14
Summary
After about 100 days on Twin Health, Victoria Song reported improved metabolic labs: all but one liver enzyme returned to normal, triglycerides fell 70 mg/dL, inflammation dropped, and insulin resistance eased.
The AI “digital twin” used continuous glucose monitor data, smart-scale and blood-pressure readings, lab work, and meal logs to predict how foods would affect her glucose and suggest changes such as walking after meals.
Twin Health paired the app with a human coach and doctor, and Song said the check-ins pushed her back into training; she recently ran an uninterrupted 5K and resumed strength work, though weight loss and medication reduction lagged.
The program remains limited to corporate or private health plans and is aimed mainly at people with metabolic issues, while the heavy data-logging burden left Song questioning how scalable digital twins are for higher-stakes medical uses.
Can an AI metabolic “digital twin” really improve fatty liver and insulin resistance, or is the real breakthrough just relentless tracking and coaching?
How much sensitive data should you trade for better health when apps like Twin Health promise personalized food predictions and measurable metabolic gains?
If a digital twin predicts your glucose well but cannot model your body realistically, should it guide treatment at all?