Difference and Bridging:Social Cognition of Human-Computer Communication
The smooth communication between ChatGPT and human users relies on its internal language model,and in terms of specific semantic processing mechanism,the communication between ChatGPT and human beings shows a kind of"predictive processing"process,in which the mutual cognition in human-computer communication is a two-way predictive coupling.However,social cognition in human-human communication does not only rely on the operation of internal mechanism,but also on the interaction among humans.Moreover,the interaction is independent and cannot be explained in the framework of individual mechanism.Cognitive results produced by the interaction are embodied,relational,and emergent,which makes the human-human communication a participatory sense-making process.However,ChatGPT's communication with human beings is not embodied,but completely relies on the operation of internal model mechanism,therefore,its communication process is only a one-way recognition of the meaning given by human beings,and what is provided to the users is only a kind of pan-situational,intellectualized and empirical response.If human-machine communication wants to achieve the effect of human communication,it is not enough to improve the machine unilaterally close to the mode of communication among humans,but also necessary to bridge the gap between humans and machines.The extensive participation of machine in the process of communicating with human beings will inevitably bring about a change in the connotation and boundaries of social cognition,so we should look for a way out for the human-machine communication in the dynamic change of the relationship between humans and machines.
NETL
NSTL
万方数据
