It Is Time to Dissolve Old Dichotomies in Order to Grasp the Whole Picture of Cognition

摘要

Models of efficient computation aiming for artificial general intelligence routinely draw a wealth of inspiration from the human brain and behavior. This applies to many diverse details and building blocks, and the most notable ones so far are artificial neural networks. As soon as it comes to more general architectural and algorithmic questions difficulties arise as there is a wide variety of models purportedly describing the basis and the working of specific mental processes. Here, it shall be sketched how a novel cognitive architecture under the name of the Ouroboros Model allows the reconciliation of many competing views by offering an overall conception, in which earlier attempts can be traced to specific and limited perspectives focusing on particular features, tasks and contexts. It is claimed that the Ouroboros Model constitutes a novel and promisingly comprehensive approach, which is still waiting exploitation for detailed formalization, modelling and working implementations.