- seCtIon FoUR: eVoLUtIon
learn from human players a function that
allows it to capture some ‘intuitive’ sense
of good board position.
So can these machines think? They
certainly have limitations as well as
strengths: they often require lots of
ancillary information and large numbers
of human examples to learn from, and
tiny perturbations can result in crucial
misclassifications (Szegedy et al., 2014).
But the answer really depends on what
you mean by ‘think’, and investigating
that, Turing argued, is not a useful way
forward.
Instead, Turing chose something alto-
gether different for his test: whether a computer could hold a conversation with a
human. Descartes had claimed this to be impossible and, interestingly, it was one
of the tricks attempted by the Turk. In its earliest version, having finished the chess,
the Turk would invite people to ask questions and reply to them by pointing at
letters on a board. But this was soon dropped from the show. Although audiences
could just about believe in a chess-playing automaton, when it claimed to be able
to answer questions they assumed it was just a trick and the fascination was lost
(Standage, 2002). Perhaps holding a conversation has always been implicitly per-
ceived as harder than playing chess.
The Turk looked like a human, but Turing did not want appearance to confuse his
test for a thinking machine, so he cleverly avoided this problem. First he described
‘the imitation game’, which was already a popular parlour game. The object of
this game is for an interrogator or judge (C) to decide which of two people is a
woman. The man (A) and the woman (B) are in another room so that C cannot
see them or hear their voices and can only communicate by asking questions and
receiving typed replies. A and B both try to reply as a woman would, so C’s skill lies
in asking the right questions.
Turing goes on:
We now ask the question, ‘What will happen when a machine takes the
part of A in this game?’ Will the interrogator decide wrongly as often
when the game is played like this as he does when the game is played
between a man and a woman? These questions replace our original, ‘Can
machines think?’
(1950, p. 434)
Turing provides a critique of his own test. He points out that it neatly separates the
intellectual and physical capacities of a person and prevents beauty or strength
from winning the day. On the other hand, it may be too heavily weighted against
the machine. A human pretending to be a machine would always fail if tested on
arithmetic, and he wonders whether this test is unfair because machines might
be able to do something that ought to be described as thinking but that is very
different from what a human does. He concludes, though, that if any machine
‘at the end of the
century the use of words
and general educated
opinion will have
altered so much that
one will be able to speak
of machines thinking
without expecting to be
contradicted’
(Turing, 1950, p. 442)
FIGURE 12.6 • Progress in artificial intelligence
has been dramatic, from Pascal’s
earliest calculating machine to
AlphaGo, shown here playing Lee
Se-Do. Does AlphaGo’s apparent
creativity suggest artificial
consciousness?