The Dexcom-Nutrisense deal is less about hardware and more about what happens after the sensor ships.
Dexcom already owns the CGM market. What they're buying is the behaviour layer: dietitians, coaching, and a subscription model that keeps users paying month to month rather than just replacing sensors.
The real tension is human vs. AI coaching. AI runs at roughly £100-500/year with 365 daily touchpoints. Human dietitians cost up to £4,000/year with maybe 24 sessions. Clinical safety favours humans; unit economics favour AI. Combining both sounds elegant until you decide who overrides whom.
Where does liability sit when an AI coach contradicts a registered dietitian's plan?