Backlinks
AI Safety — programmeBiological misuse uplift: from model answers to physical harmFraud automation: does AI change conversion, attacker cost, or successful cash-out?Live face recognition as a state-power pathway: accuracy is not the only limitCorrelated AI-service failure: what failed together, and why?Gates’s August 2026 essay: institutional adaptation rather than a credible global pauseAI-assisted AI research versus a self-sustaining recursive loopThe July 2026 OpenAI–Hugging Face incident: a bounded but real control failureAmodei’s September 2026 essay: pacing is a verification proposal, not a pause