GPT Assistants and the Challenge of Personological Functionalism

The paper discusses whether it is correct to speak of “generative artificial intelligence” – a concept that is not within the scope of AI research. The discussion suggests that it is premature to claim that humans are being replaced by GPT assistants such as ChatGPT in the field of sociocultural digital communication. Personological functionalism, which would justify the replacement of people by machines, is based on the psychofunctionalism of Ned Block, who proves the need to psychologize machine functionalism by introducing “meaning” as a criterion for passing the original Turing test. For personological functionalism, in addition to “meaning” the minimum necessary requirements of the Turing test include “creativity.” The paper shows that GPT Assistants do not pass this creativity test. To demonstrate the inability to pass a Turing test for meaningfulness, the Block machine was modified in a pair of 1978 and 1981 papers by combining the neurocomputer with symbolic versions. For the now further expanded Block test, the argumentation of previous versions is preserved and strengthened, leading to the conclusion that machines like GPT Assistants are not capable of fulfilling either the roles of psychological functionalism or personological functionalism.