Dwie metody przetwarzania informacji w głębokich sieciach neuronowych a właściwości systemów inteligencji. Próba porównania i testowania

The rapid development of AI technology in recent years has resulted, among other consequences, in renewal and dominance of connectionistic way of description of cognitive activities of both humans and computational machines. This approach results in new possibilities to compare the performance of digital AI-type computational artefacts with information processing systems of biological and evolutionary origin. The aim of this paper is to show how consequences of this approach towards the problem of computational intelligence look like. The second goal of presented research is to test the thesis of the possible superiority of systems described as immortal computing type over systems presented as mortal computing type.