Algorithmic labor (Lalgo)

Algorithmic labor, denoted Lalgo by analogy with the human labor factor, refers to the delegation to algorithms not only of execution but also of coordination and cognitive control within an organization. Theorized in economic sociology by Schildt, then by Kellogg, Valentine and Christin, this concept breaks with the classic reading that placed computing within fixed capital. With usage-based agentic AI, billed for instance per token or per API call, technology spending ceases to be a depreciated investment and becomes a variable cost strictly proportional to the task performed, exactly like piece-rate pay. Algorithmic labor then behaves as a genuine variable factor of production competing directly with human cognitive labor, for which it tends to substitute. This reclassification has profound consequences for the firm's cost structure and organization, by collapsing the marginal cost of coordination. For risk management, algorithmic labor introduces new exposures, opacity, error correlation and dependence on infrastructure providers, which form the core of the insurability questions of the agentic economy.

A firm processes thousands of audit reports in minutes by invoking usage-billed agents, mobilizing this algorithmic labor exactly as it would have hired analysts, but without any fixed payroll cost.

Lalgo, travail algorithmique, algorithmic labour, algorithmic labor