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.
the one that seems to be I wouldn't say accurate but the most accurate the less accurate is is the Apple watch and I think what they're doing since they can actually have ECG — I think they're just interpolating a bunch of data so when they see bad data they just kind of replace it with what they think the data actually should be
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.
I think they'd start build the algorith tolerates that for lack of a better word because I'll see it do that right I'll see it go from 100 to 150 and I'm like yeah get that's not even physiologically possible why yeah why wouldn't you sample that out or ask a second order question