AI researcher lays out the safety paradox: pausing algorithms while compute piles up maximizes risk
RichmanRonald · x · 2026-09-14
AI researcher kellerjordan0 argues AI safety is paradoxical: the Hugging Face incident—avoidable with better preparation—has been the most safety-promoting event of the year, showing any strategy's ultimate effect is hard to predict.
He sketches a policy that would maximize AI risk:
- Immediately pause algorithmic progress at labs while hardware-driven compute keeps piling up exponentially
- Focus intensely on safety to avoid spooking the public with industrial accidents
This minimizes accidents over five years, keeps the public and governments calm, yet leaves massive compute accumulated—so when algorithmic progress restarts, compute far outstrips safety readiness.
More from AGI Musings
- Practitioner: Huge Gap Between AI Hype and Real-World Agent Deployment — annbordetsky · 2026-09-14
- Attackers Stole METR API Key and Burned ~$600,000 in AI Credits Over Three Weeks — beffjezos · 2026-09-14
- Perry Metzger: History Has Already Proven the LLM Open-Release Camp Right Over Doomers — bradneuberg · 2026-09-14
- Debate: AI Excels at Math and Code, but Will It Ever Master Humanistic Judgment? — _onionesque · 2026-09-14
- Michael Burry: Big AI execs' 'slow down' talk is self-serving IPO hype — SumitGup · 2026-09-14
- Sam Altman on Where AI Is Headed — ResultBackground2450 · 2026-09-14