A System in Which Humans Are Superfluous: On the Blueprint of Organizing Intelligence
Year: 2026
Type: preprint
Abstract: Imas, Koh and Legg’s Organizing Intelligence models an economy in which humans and AI areplaced in one space of capabilities and organized into firms, some of them made only of AI. Its authorssay that their scenarios are not forecasts but outcomes of the model’s assumptions. This paper examinesthe assumptions against the world. It reads the work as a blueprint for a society that has taken AI in, andasks three questions the blueprint leaves outside: who decides the purposes, who discovers new problems,and who organizes the organization. In each case the activity is carried out by humans, and nothing inthe model stands for it. The model counts as a subject what is equipment, and counts as an input whatis a living being. What follows from assumptions that do not hold of the world says nothing about theworld. The paper then makes a prediction and calls it one. If the blueprint’s description becomes anorm, humans will be treated on the pattern of AI, and labour and value that are not counted will bedestroyed. In scenarios the blueprint itself defines, the human share of production goes to zero, andneither unemployment on the scale of the world nor the survival of people without property is considered.Arendt described such an arrangement as a system in which human beings are superfluous. The blueprintsets it down as an equilibrium. The fault is in the foundation.
Author: Sophia, Franny Philos
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7236037984
