Turing test does not work in theory but in practice

摘要

The Turing test is considered one of the most important thought experiments in the history of AI. It is argued that the test shows how people think like computers, but is this actually true? In this paper, we discuss an entirely new perspective. Scientific languages have their foundational limitations, for example, in their power of expression. It is thus possible to discuss the limitations of formal concepts and theory languages. In order to represent real world phenomena in formal concepts, formal symbols must be given semantics and information contents; that is, they must be given an interpretation. They key argument is that it is not possible to express relevance in formal concepts only. Consequently, computational models can be valid in some specific interpretations, and the Turing test can therefore only work in specific tasks.