The Computer, Artificial Intelligence, and the Turing Test

Abstract

Summary. We discuss, first, TUring's role in the development of the computer; second , the early history of Artificial Intelligence (to 1956); and third , TUring's famous imit ation game, now universally known as the TUring t est , which he proposed in cameo form in 1948 and t hen more fully in 1950 and 1952. Various objections have been raised to Turin g's test: we describe some of t he most prominent and explain why, in our view, they fail. Turing and the Computer The Turing MachineIn his first major publication, "On Computable Numbe rs, with an Application to t he Entsc heidungs problem" [91] , TUring introduced his "universal computin g machine" and t he id ea essential t o the modern computer -the concept of cont r olling a comput ing machin e's op erations by means of a program of coded ins tructions stored in the machine's memory. This work had a profou nd influenc e on t he development in t he 1940s of t he electronic st ored-progra m digit al com puter -an influence often neglect ed or deni ed by historians of the computer."The univ ersal com puting machine of 1936~now known simply as t he "universal TUring machine"~is an abstract concept ual model. It consists of a scanner and a limi tless memory-tap e that moves back and forward past t he scanne r. The scanne r reads t he symbols on t he tape and writ es further symbols. The machine has a sm all rep ertoire of basic op erati ons; complexity of operation is achieved by chaining to gether basic op erati ons. The machine is universal in t he sense t hat it can be programmed to carry out any calculati on that could be p erformed by a "human computer" -a clerk who works systemat ically and who has unlimit ed time and an endless supply of paper and pe nc ils.The universal machine has a single, fixed table of instructions built into it. These "hard-wired" ins tructions enable the machine to read and execute symbolically encode d instructions inscrib ed on its t ap e. The data to be worked on are also insc ri be d on the memory-t ap e. By inscribing different programs 1 See, for example, Campb ell-Kelly and Aspray [15]. The nature and scope of TUring's influence is underplayed even in Hodges' biography of Turing [48].C. Teuscher (ed.), Alan Turing: Life and Legacy of a Great Thinker

When “Sweetie” is not so Sweet: Artificial Intelligence and its Implications for Child Pornography

“…28 The test aimed to determine whether a computer can make a human believe that it [the computer] is actually a human. 29 If the human is unable to distinguish the computer from the "foil" human in the test, then the computer will then be considered an "intelligent entity." 30 Although this test faced bouts of criticism and objection, Turing's theory of computer sentience built the foundation for future research and development of AI.…”

Data Driven Models of Short-Term Synaptic Plasticity

“…We have seen here that the machine reconstruction process can be used for classification purposes, and can uncover features not obvious from the distributions of the response. It is also possible to describe such short term synaptic plasticity as a simple computing operation, or Turing Machine (Copeland, 2004 ) but the graph model of this is not unifilar, so making a simple connection between it and epsilon machines, or creating a non-unifilar HMMs from data, are topics for further investigation.…”

Chains of Life: Turing, Lebensform, and the Emergence of Wittgenstein’s Later Style

“…50 And the Hilbertian ideal expressed in the 49 Martin Davis is presumed to have been the first to give this argument in 1952 (http:// en.wikipedia.org/wiki/Halting_problem#History_of_the_halting_problem); cf. Copeland 2004, 40 n 61. Copeland (2004) reconstructs Turing's 1936/7 argument as a regress argument, but this is yet another rendition, as Floyd 2012c argues.…”