I am not at all pessimistic about expert systems. These things go in cycles and piggybag on developments. NN's were niche for a long time. Speculative hypes act

I am not at all pessimistic about expert systems. These things go in cycles and piggybag on developments. NN's were niche for a long time. Speculative hypes act as a sort of Zitterbevegung: moving fast yet going nowhere. But my suggestion is actually less ambitious: not to create confidence with reference to outside knowledge but purely on the basis of the training sample and its intrinsic statistics. [1] https://en.wikipedia.org/wiki/Zitterbewegung https://en.wikipedia.org/wiki/Zitterbewegung