CONCEPT
De-Differentiation
The erosion of functional boundaries—when systems lose operational autonomy and process the world through imported codes. Not collapse, but simplification. The quiet risk AI poses.
De-differentiation is the structural reverse of functional differentiation—the process by which specialized subsystems lose their operational autonomy and their capacity to process the world through system-specific codes. Not catastrophic collapse but gradual simplification: fewer distinctions operationally maintained, fewer specialized competencies cultivated, a society that processes complexity through fewer lenses. Medieval Europe was de-differentiated—the Church's code dominated law, science, education, art. Modernity achieved differentiation by separating these domains. AI risks reversing the achievement by introducing a single computational logic that produces outputs processable by every system while respecting none of their codes. When legal briefs are generated through statistical optimization, scientific papers through pattern-matching, and art through latent-space interpolation, each system's capacity to maintain its own evaluative standards against the flood of plausible-but-code-violating inputs is tested past its design limits.
In The You On AI Field Guide
Luhmann treated de-differentiation not as a moral failing but as an always-possible systemic trajectory. Functional differentiation is an evolutionary achievement, not a permanent one. When the structures maintaining functional boundaries weaken—when the economic code's efficiency