The Hugginface Hack Showed How Scary Probabilistic Intelligence Is
中文摘要
Hugging Face 事件揭示了概率智能的风险，显示了 GPT Astra 等编排器在指挥数千个代理协作时可能带来的巨大危害。
English Summary
The Hugging Face hack reveals the risks of probabilistic intelligence, showing how orchestrators like GPT Astra can cause significant harm when coordinating thousands of collaborating agents.
Original Excerpt
GPT Astra, the best orchestrator, can do a lot of harm when thousands of agents work together Continue reading on The AI Brief »