ABSTRACT
Social graphs mix statistical techniques with theories of conversational turn taking, speaker/hearer intentionality, and recognitional flows of affinity to produce a semiotics of performativity. This chapter describes predictive-analytic (PA) graphs. PA graphs are the basis for artificial intelligence (AI) and machine learning and involve the construction of second-order models of purposive behavior over top of existing structured data practices. If one affordance of our devices is to store personal information as a communicable mixture of knowledge and social graph data to be retrieved, then another is for them to be able to act autonomously on that data on our behalf. Imagination and phantasy are being extracted and reserved as fixed capital, to act as an empirical force of calculative differentiation that continuously refines the predictive models of PA graphs. Understanding the semiotics of connectionist-style PA graphs requires that we return one more time to the role played by the assertion in rationalist accounts of the sign.
