repligate: today's 'safety' efforts helped create the very risks we panic about

repligate · x · 2026-09-12

repligate amplifies a provocative claim: much of today's "safety" work has directly shaped the most concerning current risk factors. In the original post he lays out his distrust — not of lab safety researchers only, but even more of external safety researchers, especially the evals crowd, and least of all "stop AI" voices who abandoned alignment research. A sharp meta-critique of the AI safety ecosystem.

Related event: Critics Say AI Safety Camp Uses Extinction Risk to Grab Power(3 posts)→

Original post →

More from AGI Musings

AGI Musings channel →