The AI Labs Agreed to Slow Down. Now Someone Has to Check.
中文摘要
AI实验室同意放缓开发，将安全转化为审计挑战，且该审计范围将超出前沿实验室。
English Summary
AI labs agreed to slow development, turning safety into an audit challenge that extends beyond frontier laboratories.
Original Excerpt
Amodei’s pacing plan turns safety into an audit question, and that question does not stop at the frontier labs. Continue reading on Medium »