Searcharxiv⌕ Search

arXiv subjects

Katia Schwerzmann

Publications and source records attributed to Katia Schwerzmann.

2 recordsLinked to original sources

Toward Universal Translatability: On the Statistical Ontology of Generative AI

This paper argues that the problem of translation functions both as a point of origin and as a normative horizon for contemporary generative AI. In current discourse around ''world models,'' their attainment is framed as the achievement of internal consistency, cross-modal alignment, and physical grounding. Taken together, these demands converge on an ideal of the model as a universal translation machine capable of establishing lossless correspondences across linguistic, visual, and physical representation spaces. I show that this ideal presupposes a metaphysical commitment that I call a statistical ontology: the belief that reality has an underlying, computable statistical structure that can be recovered through sufficient data and modeling, rather than being shaped and accessed through situated human interpretation and labor. Tracing this commitment back to Warren Weaver's 1949 Memorandum on machine translation, I outline how translation was historically recast as code deciphering and, later, as the alignment of probability distributions across structured outputs. Finally, I contrast this paradigm with approaches in Translation Studies that foreground opacity, power asymmetries, and the translator's positionality. I argue that generative AI's statistical ontology is pre-Kantian, pre-critical and obscures the infrastructures and human labor that make machine translatability---as alignment---possible in the first place.

cs.CL↗

Ruled by the Representation Space: On the University's Embrace of Large Language Models

This paper explores the implications of universities' rapid adoption of large language models (LLMs) for studying, teaching, and research by analyzing the logics underpinning their representation space. It argues that by uncritically adopting LLMs, the University surrenders its autonomy to a field of heteronomy, that of generative AI, whose norms are not democratically shaped. Unlike earlier forms of rule-based AI, which sought to exclude human judgment and interpretation, generative AI's new normative rationality is explicitly based on the automation of moral judgment, valuation, and interpretation. By integrating LLMs into pedagogical and research contexts before establishing a critical framework for their use, the University subjects itself to being governed by contingent, ever-evolving, and domain-non-specific norms that structure the model's virtual representation space and thus everything it generates.

cs.CY↗