The paper introduces an unconventional way of reasoning consisting in a selection of entities called memes that inhabit Agent's Working Memory (WM). Agent's concluding belief appears as a population of identical memes that in a certain period of time dominates Working Memory. We discuss mechanisms of meme interactions, early experiments with a simulation model of meme-based Working Memory and its psychological justification, as well as potential possibility of meme-based object recognition and meme-based action planning.
展开▼