Artificial Intelligence and Software Engineering

This book is aimed at the computer-literate person who wishes to find out about the reality of exploiting
the promise of artificial intelligence in practical, maintainable software systems. It cuts through the
hype, so commonly associated with discussions of artificial intelligence, and presents the realities, both
the promise and the problems, the current state of the art, and future directions.
It is not another expert systems book. Expert systems are viewed as just one manifestation of AI in
practical software; the issues raised in this book are much broader. Expert systems are discussed, but as
a source of lessons about the dangers as well as the beneficial possibilities for adding AI to practical
software systems.
In the opening three chapters, we take a long hard look at the conventional wisdom concerning software
engineering—what the goals are, the rationale behind them, and the realities of the situation. This is
done in part because efforts to engineer Al-software appear to undermine the accepted foundations of
conventional software engineering so we need to determine how solid the foundations really are; it is
also done because in attempting to engineer Al-software we subject the standard procedures of software
design and development to close scrutiny—our attempts to build robust and reliable AI-software
provides a magnifying glass on the conventional procedures.
Chapter 4 elaborates on the prototyping schemes described in Chapter 3 and uses this well-accepted
methodological strategy to take us on into the more contentious domain of evolutionary and exploratory
software design and development. This move places us squarely within the general paradigm (i.e.
incremental system development) from whence an effective model for engineering Al software will
emerge. This chapter concludes with a presentation of the conventional paradigms for software.

Categories: ECE, CSE, COMPUTER
Download Here