The research challenge for symbolic and neural approaches

Neural state machines are known to have universal computing powers, which, through methods such as iconic learning can be turned towards cognitive tasks. This makes the division of allegiances to symbols or neurons appear somewhat false, not to say counterproductive. The states of the net, the states of the input and output are clearly symbols that are being processed by the net. The key value added by the neural nature of the approach is that the relationships between symbols can be absorbed by being in touch with the way that these occur in the artificial organism´s environment. The strategy reported in this paper is at an early stage of being proved to be feasible. There is a vast amount of theoretical and experimental work to be done against realistic problems in, say, robotics or office automation. Prospects for good mixes of conventional and neural state machine computing methods look very promising and need to be taken forward by the engineering community