You’ve already asked it something personal, haven’t you?

You typed in “Should I take a new probiotic to help with my BV symptoms?” or maybe “Why do you think I’m so tired all the time?” into your fave chatbot…

And it answered. Fast, fluent, four tidy paragraphs, sounding completely sure of itself.

About one in four American adults did some version of that in the last thirty days. So let’s stop treating it as a fringe behavior and talk about it honestly.

You probably already know I use AI every single day. And I’m really excited about the future of this technology in medicine. I’ve seen AI make connections between data points for patients that left me astounded.

But we’ve got to be realistic about its limits at this point in time, especially when it comes to health.

Feed it questions about marketing or how to lay out your living room all day long.

Don’t just blindly trust it when it comes to what to do with your body.

So, as the most AI savvy person I know, let me fully explain to you exactly why you can’t blindly trust it (especially if you’re a woman!) and how to use it properly👇

The best prompt in health is eleven words long.

“What would you need to know about me to answer this properly?”

Ask that before you ask your actual question.

Chatbots answer as though every person asking is the same person. It won’t take your history unless you make it. That prompt makes it.

What comes back is a list. Your age, your cycle pattern, what you take, what your last labs showed, what your mother’s history is. Answer the ones you can, then ask your real question. Watch how differently it responds when it’s working from your information instead of from an average.

Try it tonight on something small.

It can still be wrong, confidently, in the same tone it uses when it’s right, so treat any answer as a first draft rather than a verdict. And whatever it gives you belongs in front of a licensed human, not instead of one.

This is the mechanic underneath The Longevity Lab. Not a chatbot guessing about you, but a copilot you’ve actually taught, running on labs and tracking you already have, with my clinical framework doing the steering. Doors open next month. Click here to get on the waitlist and snag a discount!

Did you know that only around 5% of clinical trials report their results separately for women and men? Not 5% that include women. 5% that bother to say what happened to the women who were in them.

And the number I can’t shake: across a population-wide analysis spanning two decades, women were diagnosed later than men for more than 700 conditions. Four years later, on average.

I’m flagging it because it explains something you already feel, and because it’s the reason a general-purpose AI is thinner on you than on anyone else. You can’t out-Google an evidence gap.

Read the full story here →

I had a colleague recommend a book to me recently that you have to read if you haven’t yet.

It’s called All in Her Head, by Elizabeth Comen. She’s a breast cancer oncologist, now at NYU Langone’s Perlmutter Cancer Center, before that Memorial Sloan Kettering.

She goes through eleven body systems, one at a time, and lays out what medicine got wrong about women in each one, and how much of it is still sitting in the textbooks.

I’m putting it here this week because the numbers I flagged above didn’t come from nowhere. Five percent of trials reporting on women separately is the end of a very long story, and this is the book that tells it. It was a 2025 PEN/E.O. Wilson finalist, which is a science writing award, not a wellness one.

Fair warning, it will make you angry. I’d read it anyway.