As dificuldades da inteligência artificial forte o argumento de John Searle e a teoria conexionista

Abstract

“Can a machine think?”. In 1980 John Searle published the article Minds, Brains and Programs, which would change the entire debate around the so-called Strong Artificial Intelligence. In this article Searle argues that the limitations of Strong AI are outlined by its inability to causally generate intentionality. Later, Searle revisited his argument. In 1990, in an article for Scientific American, Searle responded to criticisms and counterpoints by refining his approach to the argument against Strong AI. In this article, the core of his argument against Strong Artificial Intelligence is laid out in the subtitle: “Minds are semantic, programs are syntactic”. Most of the most challenging replies against Searle's argument came from sketches of connectionist theory, especially the argument of Paul M. Churchland and Patricia S. Churchland, who considered the use of artificial neural networks for the study of human cognition as a possible solution for Searle's argument.