自我升级智能体的逻辑与认知问题

Logical and Cognitive Issues of Self-Improvement Agents

自我升级智能体的建立使人们对自我意识的研究有了一个程序化的标准,借助这种形式化的方法有可能弥合学界关于机器意识的分歧,破解机器意识研究面临的困局。但它也有逻辑上的局限。生成主义为自我升级智能体的提出奠定了认知基础。自我升级智能体的成功为生成主义提供了一个强有力的例证。尽管自我升级智能体向机器真正具有自我意识前进了一大步,但是人们只能说它具有了“功能意识”。造成机器意识困局的症结源自分析哲学传统与现象学传统的分歧和偏颇。解决的出路在于:从对立到相容,从互斥到互补,进而达到融通的新境界。

The establishment of self-improvement agents has offered a programmatic standard to the research of people’s self-awareness.With this formal method,it is possible to bridge the differences in machine consciousness in the academic world and break the predicament facing machine consciousness research.But it also has logical limitations.Enactivism lays the cognitive foundation for the proposal of self-improvement agents.The success of self-improvement agents provides a strong example of enactivism.Although the self-improvement agent has taken a big step towards the machine’s real self-awareness,one can only say that it has“functional consciousness.”The crux of the predicament of machine consciousness stems from the differences and biases between the tradition of analytic philosophy and the phenomenological tradition.The solution lies in:from opposition to compatibility,from mutual exclusion to complementarity,and then to a new realm of integration.