Robotics Reference
In-Depth Information
Bowden in his classic tome
Faster Than Thought
:
1
Aesthetic judgments, which usually involve the consideration of
the overall effects of sound or shapes, will be outside the proper
province of computing machines for many years to come. [3]
So far so good. Bowden qualified his assertion by the simple expedient of
including the words “for many years to come”. But his next sentence is an
example of an unnecessarily pessimistic forecast that is not uncommon
in relation to some of the goals of AI —a forecast that usually expresses
itself in the proclamation that “such and such will never happen”.
It seems most improbable that a machine will ever be able to give
an answer to a general question of the type: “Is this picture likely
to have been painted by Vermeer, or could van Meegeren have
done it?” It will be recalled that this question was answered confi-
dently (though incorrectly) by the art critics over a period of several
years.
As we saw in the previous chapter, this particular case of unwarranted
pessimism was put to the test exactly 50 years after Bowden wrote these
words, when a team at the University of Maastricht started work on a
promising project called Authentic, to distinguish genuine paintings by
famous artists from fakes.
2
The empirical proof that computers
can
be creative, for which several
more tangible examples are presented later in this chapter, does much to
refute the “computers can't think” attitude of some AI sceptics. But be-
fore we explore several examples of creativity by computers, there is one
specific aspect of the Ada Lovelace argument against machine creativity
that requires some comment. Her words, “It can do (only) whatever we
know how to order it to perform”, words that have been echoed countless
times since she wrote them, do not present such a clear-cut argument as
might first appear. The problem with this argument lies in the notion
of randomness. Just about every computer program operating in a cre-
ative field utilizes some form of randomness in its decision making. One
1
A ground-breaking compilation of contributions from 24 computing experts, including Alan
Turing (who wrote the chapter “Digital Computers Applied to Games”). In his preface Lord Bowden
paid tribute to Ada Lovelace, writing that “Her ideas are so modern that they have become of great
topical interest once again,...” and he includes her paper on Babbage's analytical engine as an
appendix to the topic.
2
It should also be recalled that six years before work started on the Maastricht project, another,
more dramatic example of pessimistic forecasting was permanently eradicated, when Deep Blue
disproved the dictum that “a computer will never beat the World Chess Champion”.
