CONCEPT
AI as Mashup
Lanier's reframing of large language models not as autonomous intelligences but as new kinds of social collaboration mediated by computers — mashups of uncredited human work whose apparent originality conceals the millions of specific contributions from which their capability was built.
The mashup reframing is Lanier's alternative to the dominant vocabulary of artificial intelligence. Where the industry uses verbs that attribute agency to the machine — it 'generates,' 'creates,' 'understands' — Lanier insists the accurate description is that the machine recombines, remixes, and aggregates human work performed elsewhere and earlier. The reframing is not merely rhetorical. It carries specific analytical consequences. If AI is a mashup of human work, then the appropriate question is not 'what can the machine do?' but 'whose work did the machine mash up, and what do they deserve?' If AI is a new form of social collaboration, then the appropriate frame is not philosophy of mind but economics and labor relations. The reframing restores visibility to what the standard vocabulary erases: the humans inside the machine.
In The You On AI Field Guide
Lanier articulated the mashup framing most clearly in his March 2023 essay