Robotics Reference
In-Depth Information
of no avail, and the psychoanalyst insists that he be allowed to try
his hand. The psychoanalyst finds no syndrome for which he can
account in his system. The machine has no Oedipus complex, for
it had no parents. Since it had no childhood, plunging into its past
does no good. The machine has failed to respond to any of the
traditional forms of psychotherapy.
Where do we turn now? To none other than the Therapeutic Pos-
itivist, or T.P. The T.P. sits down near the machine, asks it a few
questions and discovers that it is perplexed about the problem “Can
Machines Think?” The job of the T.P. is to show the machine that
it has been making “metamechanical” statements. In other words,
in discussing the problem with itself, it has really only been recom-
mending changes in its calculus, in the binary code.
10
This revela-
tion should make it clear to the machine that it was only tussling
with a pseudo-problem, and it will thereupon desist from making
“metamechanical” utterances. If the T.P. can perform this therapy
upon the machine, we have,
a fortiori
, shown that the machine
does
think, since it has been able to misuse its thinking powers! This, I
suggest, is the experimental crucial. We may leave it to the engineer
to work out the technical elaborations. [5]
Turing's research into thinking machines quickly advanced from the the-
oretical and the philosophical into the practical. In 1952 he “wrote” a
Chess program. Lacking a computer powerful enough to execute it, he
himself simulated the computer, taking about half an hour to calculate
each move. Only one game was recorded, in which the “program” lost to
one of Turing's colleagues. During the Turing era games became a sub-
ject of attention amongst those interested in computing, and since then
playing games of skill has consistently attracted Artificial Intelligence re-
searchers. From the mid-1950s onwards Chess in particular became a
popular measure of how much progress had been made in AI, the rea-
son being that Chess and other strategy games are generally regarded as
pursuits that require intelligence in order to play them well.
Turing never saw his ideas in Chess reach the embodiment of a com-
puter program. Within two years of his hand simulation he was dead,
at the age of 41. He had been prosecuted as a result of his homosex-
ual activities and had become viewed as a security risk by G.C.H.Q.,
the centre of Britain's post-war work on code breaking. The depression
he went through following that conviction is assumed to be the main
10
I.e., its programming.
