L’emozione del pensiero. Alan Turing e l’origine dell’intelligenza artificiale

Although much less well known than that of “artificial intelligence”, the problem of “artificial emotions” occupies a significant position both theoretically and historically. From an historical point of view, it actually already appears in Alan Turing’s 1951 article, which is one of the starting points of reflection on “artificial intelligence”. In it, Turing actually states that he intends to avoid this issue, replacing it with the famous “imitation game”. The word “thought”, in fact, does not have an exact definition: it is rather an “emotional idea” with which human beings designate verbal or symbolic processes whose functioning they do not yet understand. In order to reproduce them in a machine, it is first necessary to describe them exactly, but doing so removes the aura of processes whose functioning they do not yet understand. In order to reproduce them in a machine, it is first necessary to describe them exactly, but doing so removes the aura of humanity with which they are surrounded. Although Turing only fleetingly addresses the problem, the same criterion also applies to emotions (e.g. love). Progress in the mechanical reproduction of the manifestations of thought and emotions thus coincides with a progressive understanding of humans beings and at the same time with their disenchantment. However, Turing glimpses a possible limit in consciousness: it too has an affective meaning, but an imitation game involving it would require an infinite regress. Turing’s reflection thus implicitly poses the problem of the consistency of a human emotion that cannot be thoroughly explained.