Inside and outside the Chinese room

Abstract

In Searle's critique of what he calls strong Artificial Intelligence (AI) — a critique which consists essentially of the Chinese room gedankenexperiment — he argues „that strong AI must be false, since a human agent could instantiate the program and still not have the appropriate mental states.“ It is shown that Searle's gedankenexperiment — which rests last of all on an (incorrect) identification of knowledge in general with explicit knowledge and the (incorrect) assumption that a mental system must be transparent for itself — does not exemplify what it pretends to — namely the structure of a system of strong AI —, since the Chinese room must necessarily possess properties which are by no means properties which a computer system must have necessarily to be a language processing AI system. Therefore, the Chinese room doesn't instantiate such a system and, therefore, the gedankenexperiment doesn't tell anything about the possibility or impossibility of strong AI.