Inteligência artificial: um perguntar pelo homem?

Abstract

This thesis aims to discuss the feasibility of the AI - Artificial Intelligence - in its strong version. Our working hypothesis asks if the AI project does not run the same risk as Alchemy, that is, to aim for something considered impossible. To deal with this question we take as philosophical basis John Langshaw Austin's critique of the dichotomy, the stratified conception of Reinhart Koselleck's time, and Douglas Hofstadter's notion of a strange turn. We assume that the debate about the objectives of the IA project is strongly influenced by the dichotomous form of arrangement of ideas. Such modeling produces a deadlock in the debate, since the participants who participate do not consider the possibility that their dialogue partner can have a minimum of reason. We propose here to present a proposal that can overcome the limits imposed by the dichotomous form and maintain that the IA project is at least an admissible proposal.