Content Form:APRJA 13 Pierre Depaz
Pierre Depaz
Shaping Vectors
Discipline and Control in Word Embeddings
Abstract
This article investigates how the word embeddings at the heart of large language models are shaped into acceptable meanings. We show how such shaping follows two educational logics. The use of benchmarks to discover the capabilities of large language models exhibit similar features to Foucault’s disciplining school enclosures, while the process of reinforcement learning is framed as a modulation made explicit in Deleuze’s control societies. The consequences of this shaping into acceptable meaning is argued to result in semantic subspaces. These semantic subspaces are presented as the restricted lexical possibilities of human-machine dialogic interaction, and their consequences are discussed.
Section title
Finally, we sketched[1] out how such combination of discipline and control in shaping word embeddings can affect users. Through dialogic interaction, the user probes the spatial configurations of meaning, but the exact topology of these configurations nonetheless remains elusive, and can thus impact what can be said, and what can be imagined, a new addition to the existing challenges of linguistic expression in the era of computation.[2]
