What If Safer AI Means Governable Uncertainty?
中文摘要
AI安全可能不在于消除不确定性，而在于通过后果管理使智能系统保持可控。
English Summary
AI safety may rely on managing uncertainty to keep intelligence systems accountable through consequences rather than simply eliminating it.
原文节选
The future of AI safety may depend less on eliminating uncertainty and more on keeping intelligence systems reachable by consequence while… Continue reading on Medium »