When the humanities and AI are usually brought together, it is under the rubric of “ethical AI,” in which ethics is outsourced to the humanities. To move away from this increasingly bankrupt and problematic formulation and to help build a relationship based on the similarities between these two areas, this talk explores how fundamental concepts and axioms within critical theory and natural language processing intersect: from the notion that value stems from difference to the mapping of latent and manifest spaces. By moving from these similarities – not differences – we can begin to understand the limitations of existing generative AI models and why their insights seem to offer critical theorists, deja vu.

Webpage