AI Safety Expert: Pre-installing 'Brakes' Makes Slowing Down Scaling Less Daunting

Miles_Brundage · x · 2026-08-10

Miles Brundage discusses the impact of cautious AI safety measures on the pace of technological progress. He notes that while how much safety protocols will slow down naive max scaling and recursive self-improvement (RSI) remains an open question, having 'brakes' pre-installed makes the prospect of a slowdown far less intimidating, giving the industry a viable option to manage risks.

Related event: Ex-OpenAI Adviser Brundage Warns of AI Regulatory and Safety Blind Spots(15 posts)→

Original post →

More from AGI Musings

AGI Musings channel →