This week takes The Unsupervised Learning podcast in a somewhat different direction. In response to a common listener request, Razib takes on his first “one-man-show,” digging into his stores of knowledge of the population genetics of ancient peoples and tribes, delving into the significance of abstrusely labeled clusters like “Ancient North Eurasian” (ANE) over 60 minutes. But as anyone following this substack will anticipate, first a caveat: in these heady days of endless ancient DNA discoveries and attendant revisions to long-standing convention: everything is provisional. Razib notes that his assertions are not written in stone, as new work from researchers like Laurent Excoffier adds fresh nuance and intriguing detail to the broader evolutionary picture every few months.

This podcast takes a geographical approach, surveying Eurasian, African, Oceanian and New World populations over the last 20,000 years since the Last Glacial Maximum. Razib covers not just how populations interrelate and how they emerged, but he also touches on unique aspects of physical appearance, adaptations and natural history.

Reading:

Razib Khan: surveys of the great ancient human DNA Diasporas

Razib Khan's Unsupervised Learning