Articulating nomadic identities of radio signals

Abstract

This article presents a new materialist approach to artificial neural networks, based on experimental research in categorization of data on radio signals. Picking up on Rossi Braidotti’s nomadic theory and a number of new materialist perspectives on informatics, the article presents identification of radio signals as a process of articulating identities with data: nomadic identities that are informed by all the others, always established anew. As a resistance to the dominant understanding of data as discreet, the experiments discussed here demonstrate a way to work with a digital archive in a materialist and non-essentialist way. The output of experiments, data observatories, shows the capacity of machine learning techniques to challenge fixed dichotomies, such as human/nature, and their role in the way we think of identities. A data observatory is a navigation apparatus which can be used to orient oneself in the vast landscape of data on radio transmissions based on computable similarity. Nomadic identities render materiality of radio signals as digital information.