By Tom Leithauser
Leading AI (artificial intelligence) developers yesterday pledged to address “dangerous misuse” of their frontier models for cyber attacks and other “risks arising from growing autonomy or model misalignment, including loss of co ...