In light of the recent boom in artificial intelligence, what future and what new environments should research in media and cultural techniques focus on? This article returns to a fundamental question posed by Marcel Mauss regarding the constitution of techniques between specific and general purposes. First, it offers a brief description of the current state of machine learning and recent AI developments between specialization and universalization. Second, I propose a genealogy of artificial intelligence between nature and cultural techniques. It inquires into the historical recurrence of connectionist practices, algorithms, and styles of thinking and acting as a condition of current AI technologies. How could biological neurons become artificial neural networks that work as the infrastructural foundation of artificial intelligence? Third, what is at stake here are the notions of media and mediation themselves. Due to the naturalization and habitualization of connectionist AI, they are becoming statistically operationalized, data-intensive networks.
Published as no. 42 of the Media of Cooperation working paper series:
https://doi.org/10.25819/ubsi/11094