Statistical Intelligence and Data Capitalism, Luciana Parisi

Luciana Parisi’s presentation Statistical Intelligence and Data Capitalism discusses the arguments around machine learning AI and data capitalism from the perspective of a radical transformation in the history of machine intelligence. As she suggests, since 1980s, the primitive accumulation and valorisation of data associated with data capitalism, what Harraway calls “infomatics of domination” has encountered radical changes. The cybernetic form of capital relied no more on truth, laws and norms, but on a series of self-regulating set of procedures, organising, classifying, populations, shaping social order through the image of a network. As such, Parisi suggests, there was a total disarticulation of information from meaning – indeterminacy – replacing the soul of human with indeterminant patterns.

As contemporary capitalism tapped into communicability, what could be thought of as the expropriation of the common, which is to say language itself, a new regime of mediality emerged. The lawlessness of big data and the emergence of cybernetic capital or big data capital, introduced new political possibility around the Western metaphysical problem of ends and means. Parisi asks whether it is possible to understand in the shift in the infomatics of domination – from dominance of algorithmic patterns to evolutionary learning algorithms – an alternative vision of instrumentality that can subtract mediality from totalising image of data? What are political consequences when data has become the medium of capital? We currently cannot anticipate a finality to the system of data capture – we can no longer take dominance of data naively as only expression of automating mediality. The question remains, how to imagine an alternative when capital has become equivalent with its means of its articulation?