The Simulacra Myth of"World Simulation"—Philosophical Reflection Based on General Large Vision Models
General large vision models growing rapidly,philosophical reflections on the underlying logic of artificial intelligence technology should be made with no delays.Behind the phenomenon of'texts to videos'and'texts to images'of generative artificial intelligence is the universal spectacle of vision created by digital technology through the axiomatic logic of algorithms,which dissolves the boundary between reality and virtuality,homogeneous in essence with the power of knowledge production.As general large vision models based on pre-trained and fed data,there is a tension between the"universality"of the data and the"generalization"of the model during the process of knowledge production.The improvement of the generalization ability of the large model means that the data sources it relies on are increasingly universal and axiomatic,thus forming a public knowledge system representing digital universal rationality at the technical unconscious level.It is therefore necessary to rethink the boundary between reality and virtuality for the exploration into possible prospects of human-technique interaction in the era of artificial intelligence on the basis of focusing on different elements in technical design.
artificial intelligencetexts to video productionlarge modelsgeneralization
NETL
NSTL
万方数据
