3.3 A final caution
It may seem that I’ve been trying to refute or belittle symbolic Al to show that it is in some way discredited or simply wrongheaded. This is not so. I do not believe that Al is a failure, or that it is in any way a completed project. It is an ongoing and vital area of research that has had many successes. Like some other areas of technology, it has perhaps become the victim of these successes. Starting with early promise and much hype (most of it not generated by researchers themselves), it has failed to produce sentient androids or superbrains that keep humans as pets, but has moved modestly into the mainstream. Computer systems based on Al principles now reside quietly in washing machines, websites and wireless networks; and Al discoveries have influenced numerous other areas of computing. Predictably, the response has been pretty much that prophesied by the software scientist Bertrand Meyer:
… the well-known three-step sequence of reactions that meets the introduction of a new methodological principle: (1) ‘it’s trivial’; (2) ‘besides it won’t work’; (3) ‘anyway, that’s how I did it all along’. (The order may vary.)
Every searcher after new ideas should have a copy of this observation pinned to the head of their bed.