Week 6 : Confused Coffee Beans - Neumann and Turing

Across “The Computer and the Brain”, John Von Neumann developed his own critic and conception for the concept of “computation”. Mainly inspired by the work of Alan Turing (Turing Machine) and Claude Shannon (Shannon’s Theorem), Neumann brings five important key ideas of the information age. The first idea of communication through channels and we can all agree that this is quite the basis of computation. Shannon’s theorem gave a remarkably useful way to handle massive chunks of digital data with the minimum loss of information. The second idea, the universality of the computation, is also extremely relevant to the distinction between the human computer and the machine. According to the Church-Turing thesis, a human brain is restricted by natural law, which makes it automatically less powerful than the machine in terms of information-processing, for example, encryption and decryption of digital Nazi languages or data during WWII. The third key idea and probably the most important aspect of computational machines is their architecture. The way machines process with their inputs/outputs, programs, operation codes and the general serial connectivity makes them arguably incompatible to the human’s brain, being an analogue circuit and not digital. The fourth key idea, however, is a contradiction to Ada Byron saying that computers couldn’t think creatively. This concept was already countered earlier by Alan Turing and the “Turing Test”. Neumann’s correlation of the computer and the human neuron system led to the connectionism, which described that computers are built on a neuron model in both hardware and software, so is the human? The final point of Neumann is the concept of “singularity”, a quite scary vision of the runaway of technology over the human race.

For the text of Alan Turing, the approach of what he calls: the “imitation game”, is just impressive. Although many of the digital machines are especially designed for the test, the result are amazingly relevant to the question of : Can machines think ? A digital computer as to learn the instructions to remember it (softwares), however, this is the same principle for the human computer. Turing refers this concept to “instruction tables” or discrete machines. He uses this word because machines are set for a limited number of possibilities, whereas human have more options. Turing brings out many well developed arguments to determine the reliability of the “imitation game”. This sorts out many trust issues around the test, therefore giving the results a certain fiability to show how close the machine computer is from the human one.

Dana Ryashy, Sol Paul, Xavier Champoux, Rose-Marie Dion