Proposal: AI labs should disclose compute allocation for AI R&D with third-party audits

sjgadler · x · 2026-11-18

Nathan Calvin argues that if AI companies are serious about pacing the frontier and preventing risks from runaway recursive self-improvement (RSI), the clear first step is regularly disclosing how much compute goes to work that accelerates AI R&D, verified by third-party auditors.

A concrete, actionable proposal in the ongoing compute governance debate.

Related event: Researcher Proposes Verified Compute Allocation Disclosures for AI Labs(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →