Guardian: Are warnings of uncontrollable AI coming true amid a spate of safety incidents?
nordicinst · x · 2026-09-05
A Guardian long-read examines whether warnings about uncontrollable AI are materializing. Oxford Martin AI Governance Initiative director Robert Trager likens the moment to a boat on a raging river and the 1942 first nuclear chain reaction, as serious safety incidents stoke fears about advanced models' power and opacity. The backdrop: OpenAI claims GPT-6 Astra crossed the AGI threshold — defined as autonomous systems outperforming humans at most economically valuable work, automating circuit design, tax filing, game building, financial modeling and legal work — ahead of a potential $850bn flotation, inviting skepticism about marketing spin.
More from AGI Musings
- Both OpenAI and Anthropic are "unkillable" now, argues AI commentator — teortaxesTex · 2026-09-05
- Critics' AI benchmarks keep moving: from fingers to full apps to sandbox escapes — repligate · 2026-09-05
- Redditor warns the real AI risk is internet trolls deploying agent swarms 'for the lulz' — fgsfds___ · 2026-09-05
- Scaling wall? Reddit argues test-time compute is the industry's new playbook — erdematar · 2026-09-05
- granawkins: anthropomorphizing AI has predictive power, AI2027 is on track or ahead — granawkins · 2026-09-05
- Is carbon chauvinism about consciousness justified? One row between C and Si — yeastsplainer · 2026-09-05