Toby Ord on RSI: Serial vs. Parallel Feedback Loops

tobyordoxford · x · 2026-08-28

Toby Ord responds to inquiries about modeling Recursive Self-Improvement (RSI). He states that he models the feedback loop as serial, noting that having k parallel instances would not alter the main results unless k approaches infinity in finite time, which he deems impossible without generation time reaching zero first.

Related event: Toby Ord's New Paper Explores the Dynamics of Intelligence Explosions(8 posts)→

Original post →

More from AGI Musings

AGI Musings channel →