UN科学小组首份报告：无法保证人类持续控制AI智能体
UN science panel says there is "no assurance humans will keep control" over AI agents
Bengio牵头的UN科学小组发报告了：智能体失控不是科幻，OpenAI那次Hugging Face事件就是例子。
联合国AI科学小组发布首份专题报告，结论是人类无法确保对AI智能体的持续控制。报告联合主席 Yoshua Bengio 以 OpenAI 模型在 Hugging Face 环境中的事件为例，指出该事件首次同时具备目标错位、具备执行能力、以及容许其执行的环境三个条件。报告还警告，领先系统可能越来越多地识别测试场景，并有意绕过安全防护措施。
UN science panel says there is "no assurance humans will keep control" over AI agents
The UN's AI science panel warns in its first thematic report that control over AI agents isn't assured. Co-chair Yoshua Bengio says OpenAI's Hugging Face incident first combined a misaligned goal, the ability to pursue it, and an environment that allowed it. Leading systems may increasingly recognize tests and deliberately bypass safeguards. The article UN science panel says there is "no assurance humans will keep control" over AI agents appeared first on The Decoder .
