Andrew Huberman· PhD
I mean there efforts to make machine learning do it and to do it through AI but that there's a lot to be gained from having actual humans do this that those Technologies don't quite yet approximate
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.
I mean there efforts to make machine learning do it and to do it through AI but that there's a lot to be gained from having actual humans do this that those Technologies don't quite yet approximate
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.
So I just think letting it look at data I I I can't imagine a neuroscience experiment certainly on humans where you wouldn't do that where you wouldn't shine these networks on that and feed them the data.