Google researcher ibab argues LLM pretraining distills humanity's unpaid collective work, and charity can't settle the debt

ibab · x · 2026-11-20

Researcher ibab lays out an under-discussed ethical framing of LLMs: pretraining distills vast amounts of humanity's collective work into model weights, yet when that content was created and released as a common good, nobody knew LLMs would become possible. In a follow-up, he argues charitable giving can't resolve this dilemma, since inefficiently spending such vast profits on pet causes is fundamentally at odds with the democratic spirit. The thread pushes the pretraining-data debate beyond copyright toward how labs owe an unliquidated debt to everyone.

Related event: Google Researcher Argues Open Weights Should Return Distilled Human Commons(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →