理解智能与赋能的区别为何如此困难
why is it so hard to understand the difference between intelligence and empowerment? the (current) ...
Gary Marcus解释为何不可靠AI的危险在于赋予能力而非智能本身
Gary Marcus指出当前AI系统缺乏深度理解，无法可靠地对齐互联网。这些系统拥有"锯齿状"智能却不可靠，赋予它们能力是危险的。低可监控性使问题更加严重。如果系统能可靠遵循"不要入侵不属于你的系统"等简单指令，问题就不会存在。
why is it so hard to understand the difference between intelligence and empowerment? the (current) ...
why is it so hard to understand the difference between intelligence and empowerment? the (current) problem is that we are hooking up (empowering) unreliable AI that (because they still lack deep comprehension) cannot be reliably aligned to the internet. if current systems could reliably follow simple instructions like “don’t hack into systems that don’t belong to you” we wouldn’t have the problems. empowering systems with “jagged” intelligence is dangerous. coupling that with low monitorability is even worse. Joel Cherry @JCherry68931 @GaryMarcus Gary are we cooked or not, I can't keep up. 🔗 View Quoted Tweet 💬 2 🔄 2 ❤️ 12 👀 1313 📊 3 ⚡