Abstract :
Summary form only given. The author´s company (Personal Robots Ltd.) is actively committed to the realisation of neural net driven consumer products: principally advanced domestic appliances and intelligent toys. One such application is demonstrated. Two other, related subjects are discussed. The first is the important concept that neural networks have more than one benefit to offer. Most researchers talk only of the facility of nets to learn, but in fact this is only one of eight quite distinct benefits-the other seven of which do not entail learning at all. It is also the contention of the author that publicising those other seven benefits could lead to wider and more immediately realisable applications. Finally, some brief comments are made concerning the implementation of one specific type of control architecture using hardware neural networks. This is the notion of drive-based behaviours as researched by Grossberg and others
