Being Legible to the Machine:

Algorithmic Monitoring, Energy Poverty, and Sociotechnical Justice in Olacapato (Salta, Argentina)

Abstract

This article examines the transition toward algorithmic monitoring of energy consumption in contexts of poverty, focusing on the situated case study of Olacapato, in the Puna region of Salta, during 2024. Methodologically, the study adopts a qualitative-critical approach that uses the 2024 Energy Census as both an empirical source and an analytical proxy. The survey was based on a sample of 30 households—covering approximately 75% of the estimated total of 40 dwellings—and involved structured face-to-face interviews conducted during December. The findings reveal that standardized metering systems render invisible survival practices such as voluntary disconnection, the combined use of diesel, gas cylinders, and firewood, as well as community-based energy networks. Algorithms operate as social agents that classify these strategies as anomalies or fraud,
thereby enabling punitive interventions. Drawing on authors such as Tarleton Gillespie (2016), Cathy O’Neil (2016), and Safiya Umoja Noble (2018), the article argues that algorithms function as infrastructures of power. Finally, it proposes a community-based energy audit grounded in situated metrics as a means of contesting energy governance and advancing toward sociotechnical justice