Computerphile explains GPT-3's scaling and emergent arithmetic
Computerphile · Explainer · 26 min · 2020 · YouTube title: GPT3: An Even Bigger Language Model - Computerphile
Why it's here
HN commenters cite it as a concise explanation of GPT-3's scaling jump and surprising abilities at its release.
What it covers
Rob Miles discusses GPT-3 on Computerphile, covering how scaling parameters kept improving results over GPT-2 and how the model nearly learned arithmetic without anyone knowing exactly how.