Máquinas pensantes: algoritmos de inteligência artificial como discurso e produção algorítmica no capitalismo

Abstract

The emergence of artificial intelligence and digital and automation technologies has led to many analyses of their social impacts, especially those related to the elimination of jobs, the precariousness of work, the restructuring of production, and the transformation of capitalist relations; but also regarding surveillance and control without transparency, directed behavioral changes, and the weakening of democracies. The academic debate in the fields of computing and engineering, but not limited to these, as well as media discourse, has been reflecting on algorithms as endowed with qualities of human intelligence, creativity, and even consciousness, often naturalized as neutral, in which AI algorithms are humanized as the human is biologized and displaced from its historical determinations. The research has assumed algorithms as discourse and device of power; more precisely, this study analyzed algorithms in their components and functioning, as discursive formations that integrate the capitalist mode of production. We analyze the algorithmic nature of work in the digital economy in the production of value and how the lack of analysis of this condition has led to difficulties in evaluating immaterial labor in the current mode of capital accumulation and generated so much perplexity. In addition to its practical applications and the problem of data bias, the analysis of algorithmic AI discourse allows us to understand AI-based digital technologies as an algorithmic machine that operates at the heart of contemporary power relations and transforms production relations while maintaining the exploitation of labor in the process of production and capital accumulation.