You are right, but artificial intelligence, deep learning and machine learning all use words related to human activities and the roots of machine learning perceptron, neural networks, feedback and other terms are also imitations or analogies of terms for human activities, in the end the goal of deep learning is learning patterns that are not available for neural networks with few layers. The insight is that the human mind and perceptive organs feed the mind with structures that are similar to layers. Your retina see patterns that are hard coded and are th result of evolution. Perhaps I am not selling snake oil.