人工智能與認識論有著獨特而內在的關系,使得兩者之間可以進行哲學上的互釋:一方面是對人工智能的認識論闡釋,包括揭示人工智能的認識論根基,尤其是不同人工智能綱領或范式(符號主義、聯結主義和行為主義)的哲學認知觀,以及它們進行智能(認知)模擬時與人的認知之間所形成的同理、同構、同行、同情的不同關系;另一方面是對認識論進行基于人工智能視角的闡釋,包括依托人工智能范式所進行的認知分型(推算認知、學習認知、行為認知和本能認知),進而揭示這些分型之間的多重關系。在此基礎上,還可以對人工智能和認識論之間進行動態互釋,揭示兩者之間難易互逆的關系,由此對人工智能發展走向形成有根據的預判,有助于正視人類智能和人工智能之間的互補,進而推進不同算法和認知類型的融合,并印證人們對認知本質相關闡釋的合理性。

The unique internal relationship between artificial intelligence(AI)and epistemology enables the two to have an interactive philosophical interpretation.On the one hand,the epistemological interpretation of AI includes revealing its epistemological foundation,especially the epistemological views of different AI programs or paradigms(symbolism,connectionism and behaviorism),together with different relationships featuring shared principles and structures as well as parallels and sympathies built up through the conduct of intelligent(cognitive)simulation.On the other,it provides epistemological interpretations from the perspective of AI,including cognitive classification(inferential cognition,learning cognition,behavioral cognition and instinctive cognition)based on an AI paradigm,thereby revealing the multiple relationships among these classifications.On this basis,a dynamic interactive interpretation between AI and epistemology can be carried out that reveals the degree of difficulty of their reciprocal relationships,thus providing a wellgrounded prediction of AI developmental trends, helping to confront the complementarity between human intelligence and AI,further promoting the integration of different algorithms and cognitive types,and confirming the rationality of the interpretation of the nature of cognition.