🎓 Hippocrates Meets AI 🌐
EPISODE #5: "Everything Flows"
🙋♂️🌊 A new season begins, and we're back with a bang — with what is, so far, my favorite episode of the series.
☀️ Summer 1984, Fremantle, Australia.
A 32-year-old gastroenterologist, Barry Marshall, asks his lab technician to prepare him a "drink". A culture of Helicobacter pylori, suspended in broth.
🥤 He drinks it in one go. He doesn't even tell his wife.
🤢 3 days later: foul-smelling breath (first noticed by a relative of his). Then nausea and vomiting.
🔬 10 days later, endoscopy: gastritis — the histological examination teeming with the same bacterium.
😤 Years earlier, he had presented his theory at a conference: that peptic ulcers are not caused solely by stress and spicy food, but by a microorganism. They laughed at him. "The stomach is too acidic for any bacterium to survive there," they told him.
🏆 2005: Nobel Prize.
❗ He had to drink bacteria before they believed him.
📉 But we don't need to go back to Australia in '84.
🥛 For decades, the recommendation for patients with dyslipidemia was clear: low-fat dairy. Full-fat milk, regular yoghurt, and most cheeses: in moderation — or not at all.
📅 2026: The official US dietary guidelines are revised. They now explicitly recommend full-fat dairy products. The scientific basis: newer data show a neutral relationship with cardiovascular risk.
🔄 Something I had been saying with confidence for years, revised right before my eyes.
💊 Unlearning in Medicine is difficult — but necessary.
👴 Older generations of physicians grew up with a reflex: faced with fever, cough, sore throat — prescribe an antibiotic, "just to be safe." Even in viral infections, where it does absolutely nothing.
🦠 Today we know: every unnecessary prescription fuels antimicrobial resistance — a problem inherited by future generations. The reflex is hard to extinguish, even when you know better.
🩺 And the reverse holds true on the patient side. Older generations grew up with the logic of "I go to the doctor when something hurts." The idea of having a check-up when you feel perfectly well, "just to be sure," strikes them as unnecessary — perhaps even suspicious. It is hard to convince them that preventive medicine today detects what, in the past, was discovered far too late.
⚖️ Two opposing habits that need to be "unlearned" — and both point to the same truth: letting go of what you already knew with certainty is harder than simply learning something new.
📜 We often say: " Τα πάντα ρει " — "Everything flows" — attributed to Heraclitus.
⚠️ In reality, he may never have written it in quite those words. What survives as genuinely his is something harder:
"On those who step into the same rivers, ever different waters flow."
And indeed, it is not only the river that changes. We ourselves change, each time we step back in.
🌊 Even the very phrase about flux — flows. It changed along the way, from generation to generation, until it reached us, already altered.
🧊 Something analogous happens with the artificial intelligence tools I use in this series. They too have a kind of "expiration date" on their knowledge — a moment at which their training stops, while the world continues to flow. If you had asked such a system just before January 2026, it would still have answered you with yesterday's certainty about dietary fats. A river, photographed at a single moment, hung on the wall as though it were still that same river.
🎓 In medical school, we were taught quite a few axioms. Later, I came to realize that axioms melt. In postgraduate studies, we were told categorically: "Burn your books!"
💡 The most honest physician is the one who never falls in love with what they know — because they understand that, at some point, they will need to let it go.
🌊 Thousands of years after Heraclitus, the river of medical knowledge continues to flow. And we flow with it, stepping back each time to water that is never the same.