人工智能与人类智能——从认知科学五个层级的理论看人机大战

Artificial Intelligence and Human Intelligence:AView on Human-Computer Competition from the Five-Level Theory of Cognitive Science

人类心智和认知分为神经、心理、语言、思维、文化五个层级。在人类认知的所有五个层级上,人工智能都是在模仿人类智能;人工智能是在不断进步的,但在总体上并未超过人类智能。在语言、思维和文化层级上,即在高阶认知层级上,目前人工智能都远逊于人类智能。事实上,人工智能和人类智能这两种智能方式是截然不同的。机器学习也只是对人类认知能力的一种模仿,不能作过高的评价,不必惊慌,更不能人为地制造恐慌。著名的塞尔人工智能理论"中文房间论证"的模型和标准并未过时,根据这个论证,数字计算机(digital computer)永远不会具有人类智能,因此,在可预见的未来也不会出现超过人类智能的人工智能。

The present authors hold that human mind and cognition can be viewed at five levels, i. e. , at the levels of neuro, psychology, language, thinking and culture. This paper argues that artificial intelligence （ AI, for short） simulates human mind at all the five levels of human cognition. On the whole, AI has not surpassed human mind although it has kept making progress. So far, AI still lags far behind human intelligence at the higher-levels of cognition, that is, at the levels of language, thinking and culture. In fact, AI and human intelligence are two entirely different modes of intelligence. Machine learning is only a simulation of human cognitive ability. It is unnecessary for us to overestimate it or to be scared of it. What is more, we can by no means artificially create panic. Chinese Room Argument, a famous AI model and standard put forward by John R. Searle, is not out of date. According to this argument, digital computer will never have human intelligence. Therefore, no artificial intelligence that may surpass human intelligence will appear in the foreseeable future.