Semiotics of the black box: on the rhetorics of algorithmic images

Abstract

This article critically analyses the semiotic pathways through which the new aura of algorithmic images is constructed, an aura which stems not so much from what they represent or from how it is represented but from the halo of mystery surrounding the very productive genesis of such images. Even their creators, from their super-technological laboratories, claim that they cannot fully grasp their emergence from artificial intelligence. Analysing these statements in depth, as well as the attempts that these same laboratories conduct to ‘unravel’ the mystery of the algorithmic images that they themselves fabricate and disseminate, however, one is seized with the suspicion that this mystery and aura are not due to intrinsic technical causes, but rather to the particular socio-rhetorical context in which digital and technological frontier knowledge is produced today, especially in relation to artificial intelligence. The ‘black box’ so often evoked to translate the inexplicability of artificial intelligence visual products might therefore be nothing more than a rhetorical device to protect and enhance the real black box, that of productive and industrial secrecy. In this whole process of algorithmic construction of the aura, then, the rhetoric of the unknowable image intercepts and highjacks a very longstanding trend in human cultures, in which images are precisely delegated the semiotic task of circulating the sense of a mysterious, ungraspable and unfathomable meaning.