emollick· @emollick · X·· 6 小时前AI 评分
基准测试与复杂长任务的差距
我还应指出，几乎所有基准测试都是单答案、短周期任务，但大量有经济价值的工作是复杂的长周期任务。两者都有空间，但我认为我们会发现"足够好"的智能可能并不总是够用。
I should also point out that almost all benchmarks are of single-answer short-horizon tasks, but a large amount of economically valuable work is complex long-horizon tasks. Again, there is room for both, but I think we will find "good enough" intelligence may not always be.
来源：emollick · x.com