인류는 다가오는 지능 폭발에 대한 준비가 되어 있지 않다

The emergence of recursive self-improving AI systems poses an existential risk to humanity, yet safety investments remain vastly inadequate compared to development spending. With leading AI researchers estimating a 10-50% probability of catastrophic outcomes—far exceeding nuclear safety thresholds—urgent US-China coordination and verification-based international governance frameworks are essential to prevent irreversible loss of human control over superintelligent systems.