08/02/2026
People aren't just searching for health information anymore.
They're asking AI to interpret it for them.
And that's a bigger shift than it sounds.
Many people ask AI questions like:
• "How do I lose my cortisol belly?"
• "How do I treat my histamine intolerance?"
• "What supplements should I take for my perimenopause brain fog?"
While they may seem simple, we’re increasingly asking AI some of the hardest questions in healthcare - often without realizing what information it's drawing from or how it arrived at its answer. And often, we're asking those questions after our own thinking has already been shaped by algorithms, influencers, podcasts, social media, and previous searches. That combination worries me. Not because AI can't be useful. It absolutely can.
But because the first question we ask matters.
There's a big difference between asking:
"How do I lose my cortisol belly?” and
"What are the most likely explanations for my belly, and what evidence supports each one?"
One begins with a conclusion – that may be wrong. The other begins with curiosity.
AI is often only as good as the question we ask it.
When I read a journal article, I know where it came from.
When I read a random person’s blog, I know where it came from.
When a patient brings me an AI-generated answer, I often don't know what combination of sources produced it - or how the AI weighed them.
That lack of transparency makes evaluation much harder.
As clinicians, I wonder if our role is changing.
We often hear that AI will make clinicians less important. But, I wonder if the opposite is true.
Not because patients can't access information. But because people still needs to help them recognize when a question is unintentionally narrowing their thinking.
I'm exploring this idea in an upcoming presentation on AI, misinformation, and clinical reasoning, and I'd love to hear what others are seeing.
