If you are aware of all the possibilities and dangers that arise with
the use of flexible (in other words: pliable) software, you may wonder:
are we acting wisely? Do programmed machines pose a direct or indirect
threat to mankind? Has everything been thoroughly tested? The answers
are simple: no, yes, no! The whole profession of software development
is still in its infancy and error-free software is still a utopian dream.
Even sufficient testing of traditional software still appears to be
a problem.
If we consider a few new developments in the field of information technology,
such as object orientation, knowledge technology and parallel processing,
the process of testing becomes even more difficult. An object is in
fact an entity with its own memory and its own behaviour, which can
react to a great variety of external impulses. Go ahead and test that
thoroughly. A knowledge system consists of a set of rules
of which the outcome cannot even be predicted in advance, let alone
be tested.
Knowledge technology is becoming more and more of a substitute for
our own thinking. When testing it, we are in fact encountering the limitations
of our own thinking.
A combination of the three technologies listed above, in other words,
a set of objects, each fitted with knowledge technology and able to
work parallel, will constitute a gigantic step forward in automation.
We will, however, get a piece of software that can hardly be tested
anymore, and that will be able to solve problems with a complexity and
a speed that is far beyond the grasp of our own thinking. What would
our colleague Van Vliet recommend when he summons us as I mentioned
in my introduction to exercise restraint as long as we are insufficiently
in control of technology?
Our fate as human beings is in the hands of the gods, especially in
those of the god of mechanical thinking. There is only one way out for
us, and that is to transcend our own thinking by finally waking up.