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.
This paper, we will make available by a link in the show note captions, but basically what this paper shows, is that when trying to learn something new, you want to make the difficulty of what you're trying to learn, such that, you are getting things right about 85% of the time.
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.