Send message to Author

Short abstract:

This contribution aims to discuss the consequences of a non-anthropocentric reading of cognition in the time of brain-inspired AI and its implications for human autonomy. This will be accomplished through a three-step critical exercise that seeks to reconceptualize human-machine entanglement.

Long abstract:

A renewed convergence of computer engineering and neuroscience is unfolding, facilitating the integration of bio-inspired principles into both software and hardware. Unlike previous ‘’symbolic’’ architectures, 1) brain-inspired AI systems capitalize on multi-layered and open-ended computation, making them more adaptable to nowadays informational complexity. Moreover, 2) the social space and the environment are being transformed as a function of technical objects that rely on these systems, to enable them to better operate. This contribution will first explain how these two conditions are crucial to understanding the increased agential capacity of these technologies.

These changes necessitate viewing AI as material force embodied into the decision-making system. It will then be discussed the limits of an anthropocentric idea of cognition and intelligence, which involves 1) reading AI as a form of «infrastructural cognition», understood as a network of planetary computation and 2) imaging the social space as «cognitive ecosystem», intended as an assemblage of data flows, non-human forms of intelligence, institutional and intellectual structures, and connected technologies. This entails a onto-epistemological reframe of the relationship between human and technology: no longer simple mediation and interaction but radical entanglement.

Finally, moving to the normative level, this contribution will ask: how should it be remodeled a category as central as that of human autonomy, which always subsume a form of cognitive independence? In this third step, it will draw on new materialist approaches, especially in their interpretation of non-human agency and relational assemblages – also trying to elucidate the conceivable constraints of this non-essentialist interpretation.