Andrew Huberman· PhD
For the real world, reliability beats validity as much as we can. For a lot of the things we're talking about, especially for using it as a metric of did I get better, as long as that tool is reliable.
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.
For the real world, reliability beats validity as much as we can. For a lot of the things we're talking about, especially for using it as a metric of did I get better, as long as that tool is reliable.
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.