Models are used as orders of magnitude cheaper substitutes of real thing in learning, predicting, and so on, known as «modelling».
AI, in it current state, is not good enough to serve as substitute for human, or human brain, but good enough to serve as substitute of human level intelligence. At this point, we are able to model brain of a fly.
It looks like you are confused by similarity of scheme, model, and similarity.
A model needs a map to transfer knowledge in both directions: from a real thing to a model, and from the model to the real thing, while in scheme, knowledge is transferred in one direction: from a real thing to a scheme. They toy humanoid robot is just a schematic representation of a real human.
Moreover, similar things are not models of each other. Apes are not models of humans and vice versa.
It just depends on your definition of a model, but to me, a neural network is modeled after how a human brain works.
If Apes were man-made, I would also count them as a model of a human.
AI, in it current state, is not good enough to serve as substitute for human, or human brain, but good enough to serve as substitute of human level intelligence. At this point, we are able to model brain of a fly.
It looks like you are confused by similarity of scheme, model, and similarity.
A model needs a map to transfer knowledge in both directions: from a real thing to a model, and from the model to the real thing, while in scheme, knowledge is transferred in one direction: from a real thing to a scheme. They toy humanoid robot is just a schematic representation of a real human.
Moreover, similar things are not models of each other. Apes are not models of humans and vice versa.