Interesting paper, but I would like to clarify a couple of things. First, Turing’s theory of the computer was published in 1937 and so preceded his work of cracking the Enigma code.
Second, in 1948 Turing wrote a paper on Machine Intelligence (it was not published until the 1970s because the British Government sat on it), that argues the case for what we would now call a neural network. In paricular Turing wrote that the problem with machines is that they break, or halt when they make a mistake. By contrast, humans learn by making mistakes, he said:
‘In other words if a machine is expected to be infallible, it cannot also be intelligent.’
He therefore argued for a machine that could copy some of the wiring of the brain.