CONCEPT
Cognitive and Informational Externalities of AI
The class of costs generated by AI deployment that fall on parties outside the transaction producing them — cognitive capacity eroded at scale, knowledge ecosystems degraded by the models that depend on them, worker health absorbed by individuals whose employers captured the gains. The AI economy's smoke: invisible, mounting, unpriced.
An externality is a cost or benefit that falls on parties not involved in the transaction that produced it. Environmental pollution is the canonical example: a factory produces widgets priced to the buyer and smoke imposed on the community. The smoke does not appear on the factory's balance sheet, does not factor into the price of the widgets, and persists until institutional intervention forces the producer to internalize it. Stiglitz's career-long engagement with externalities has demonstrated that the gap between private cost and social cost is where markets produce their most destructive outcomes. The AI economy produces externalities novel in kind but familiar in structure: cognitive erosion at civilizational scale, informational ecosystem degradation through recursive training, and work intensification whose health consequences workers absorb individually while employers capture productivity gains aggregately.
In The You On AI Field Guide
The first externality class is