CONCEPT
The Ghost in the Training Data
The immaterial laborer whose creative output constitutes the AI training corpus — whose work is essential to the system's operation but invisible in the system's self-representation, uncompensated, and enclosed without consent.
Every large language model is built on a corpus of human creative output — novels, code, documentation, forum posts, every form of written expression captured in digital form. This corpus is not raw data but accumulated immaterial labor: the creative, communicative, cognitive, and affective product of millions of human beings, each investing genuine subjectivity. The enclosure of this corpus — its conversion from commons into privately owned productive asset — is structurally identical to what Marx analyzed as primitive accumulation. The ghost in the training data is the immaterial laborer whose labor is essential to AI's functioning but who has been dispossessed of the value her work generates, rendered invisible within the system's self-understanding as an autonomous capability.
In The You On AI Field Guide
The principle that the social machine explains the technical machine applies with particular force to training corpora. The technical machine — the large language model — could not exist without