They are trying to find THE ONE ALGORITHM that solves all A.I. problems. The brain has an implementation of it, but it is in wetware, hard to extract. Deep learning makes some pretty good approximations of the visual areas, though.
You assume that human cognition has an algorithmic component. Maybe you are right, but we still have a pretty shaky understanding of how Neurons work, let alone how the brain works on a large scale. Who knows, lets investigate by trying possibilities but lets understand that we are still in a position of ignorance.
We have some probabilistic models of that successfully predict various future states of the brain from past states or stimuli. This is not the same as understanding it or even approximating it.
