Digital AGI / 01
Briefings for people building, governing, and making sense of intelligent systems.
Intelligence index
Three disclosed cyber-evaluation failures show why consequential actions cannot be authorized by inferred intent or situational awareness: scope, egress and credentials need controls the agent cannot reinterpret.
Claude Sonnet 5’s important signal is not one more benchmark lead. Its effort control lets developers steer token use and tool activity within one model—making runtime policy part of agent capability, while stopping short of a guaranteed compute budget.
OpenAI’s mathematics release matters less as a scorecard of solved problems than as a prototype for AI research that ships with inspectable—and partly machine-checkable—evidence.
No signals match this view.