intra-active machine learning

Scurto, 2024
reflexive essay,
new materialism, computer science

A theoretical reflection on intra-action and machine learning. Intra-action greatly contrasts with the traditional concept of interaction, which presupposes the pre-existence of materials before they meet, and a separation between objects and living subjects. Intra-action presupposes that all beings or bodies are constantly exchanging and working inseparably, into one fluid and ever-evolving phenomenon. From this intra-active entanglement emerges an ethos of responsibility with/in the world: we must matter for our matter as much as that of the Other. I argue that intra-action can help promote socio-technical practices with/in machine learning, from attending to the always already fluid roles of developers and other stakeholders in AI, and responding to the material attributes of machine learning—including uncertainties and difference-making.

Year
2024
Publications
Paper at Full Stack Feminism (2024)