The difference between mathematical neural networks and humans is how they're trained and how they function, which is not just a difference in kind. The way tha

The difference between mathematical neural networks and humans is how they're trained and how they function, which is not just a difference in kind. The way that they adjust weights and perform inference is different enough that them being "theoretically isomorphic" is not a viable argument.