Perplexity Health

I've posted here about Anthropic's Claudes (Code, Cowork, and Chat) and about the new Gemini 3.1 Mac native app. The third tool I use every day is Perplexity, and I haven't said much about it here yet.

I was an early adopter. The in-line citations felt like what Google used to be. Is Perplexity ethically perfect? That's ambiguous. Has it been invaluable to me? Yes.

This past week I've been running Comet as my only browser. Chromium underneath, Perplexity baked in. Pinned tabs, extensions, all the browser stuff I'm used to, plus a sidebar that actually understands the page I'm on. When agentic coding tools like OpenClaw started dominating the conversation, Perplexity already had their answer: Perplexity Computer. And it rocks. Now they're entering health.

Perplexity Health connects your medical records (via b.well, HIPAA-compliant), your labs, and your wearables. Apple Health, Fitbit, Withings, and Oura coming soon. Eric Topol, who wrote "Deep Medicine" in 2019 when almost nobody was talking seriously about AI in healthcare, sits on the advisory board. Data is encrypted, not used for training, not sold, and you can disconnect or delete it anytime.

So: am I ready for AI to be my concierge doctor?

Perplexity is explicit that the product is informational. Not diagnostic, not a substitute for a clinician, not adjusting medication dosages. That's the right line.

But the question I keep turning over is bigger than the privacy policy. When a tool this good at synthesis meets data this personal, what I want to know is whether the company behind it has the judgment to say no to the things that would grow the product and potentially compromise privacy and health. I'll keep using Perplexity (maybe not for Health... although... it's me, so maybe yes). I'll also keep watching the choices they make, and the ones they don't.