Jerry Tworek, CEO and co-founder of Core Automation and former VP of RL at OpenAI, observes low bias high variance methods scale better than high bias low variance ones despite lower efficiency

— You Jiacheng reposted the observation without added comment.

Original post

OPJT#126Jerry Tworek|@MILLIONINT

Low bias high variance methods are usually not very compute efficient but they are scalable High bias low variance methods have higher efficiency but hit limits of scaling quickly One of those sides has had better time historically

6:56 AM · May 20, 2026 View on X

Reposted by

YJ#826|@YOUJIACHENG