Millions Discuss Their Health With ChatGPT Every Day
ChatGPT is making a direct play for your most “sticky” context - your health.
OpenAI just announced ChatGPT Health - a dedicated space inside ChatGPT where you can connect medical records and wellness apps (e.g. Apple Health, MyFitnessPal, Function, etc.) so conversations are grounded in your real history and patterns.
They’re also designing it as a separate compartment - “Health lives in its own space...stored separately”, with separate memories, and they say Health conversations aren’t used to train foundation models. To connect U.S. medical records, OpenAI says it partners with b.well and you can remove access in Settings.
My view is that this puts downward pressure on the Portability dial.
“...based on our de-identified analysis of conversations, over 230 million people globally ask health and wellness related questions on ChatGPT every week.” - Google
Once your “life interface” accumulates inside a single system (longitudinal lab trends, appointment prep, insurance context, wearable patterns, and the conversational scaffolding you build around it) the switching costs jump. Even with good controls, the default gravity is lock-in - the more your health context becomes native to one agent, the harder it is to port your Cocoon elsewhere without losing continuity.
> The interface layer is thickening. If you disagree with my interpretation, or you’ve spotted a better signal then reply and tell me.