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.
it feels like we should be putting a lot of our energy into figuring how to make AI help us not die and have ai be instrumental in medicine and you know like if it can do all your panels and look at all your markers and be able to diagnose something with a lot of accuracy where a doctor might have a bias or whatever that feels like a very good use of AI
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.