Inteligância artificial: manipulação simbólica e intencionalidade em J.R. Searle

Abstract

This work investigates the philosophical problems arising from the concept of intentionality in Artificial Intelligence, as formulated by J. R. Searle. We first try to show that symbolic manipulation, the basis of digital computer behaviour, does not imply any evidence of understanding. We then change from a behavioural point of view, which gives the discussion the utilitarian character of computer science, for which the final behaviour of an artificial inteligence system is proof of its inteligence, to the onepresented by Searle, for whom the observation of a system's behaviour, lacking in intentionality, cannot be such a criterion. We develop this critique introducing the semantic aspects of intentionality and the problems in formalizing it through sjonbolic manipulation. We conclude by considering the philosophical problem of Artificial Intelligence as a non-empiric question.