Peter Attia· MD
people don't go big enough give you a 90% confidence interval and you'd be better off going like well I know there's 10 you know - like a million and instead they go much narrower and they end up missing almost all of them
Loading…
The headline is broadly defensible, but the qualifications matter. Effect sizes vary by population, the strongest claims rest on shorter trials, and credible voices push back on how it's typically framed.
people don't go big enough give you a 90% confidence interval and you'd be better off going like well I know there's 10 you know - like a million and instead they go much narrower and they end up missing almost all of them
Every Sunday: the week’s new conflicts and verdict changes — and nothing else.
Native comments, Twitter mentions, and Reddit threads about this claim — surfaced together so the conversation isn't fragmented across platforms.
Bookmarking — the dossier-vs-overview split is the right call. Most of the time I want overview; sometimes I want receipts.
Would love a "what would change this verdict" RSS feed. Sign me up if it exists.