Loading...
A large language model is structured like the unconscious: the (ordinary) perverse psychosis of AI
Authors
Editors
Other contributors
Affiliation
Epub Date
Issue Date
2025-11-11
Submitted date
Subjects
Alternative
Abstract
This article asks whether discursive relations with large language models (LLMs) might be able to generate new nonhuman forms of thought or subjectivity. It bridges the existing Lacanian analyses of AI which either discuss a hypothetical conscious form of nonhuman subjectivity, or address our contemporary interactions with LLMs, by considering how these existing interactions both replicate perverse human subjectivity and indicate something like an (ordinary) psychotic nonhuman subjectivity. These interactions are perverse in the sense that the LLM disavows its programmed partiality via an ostensible nonhuman objectivity which allows it to (re)present the ideas of the ruling class – the fantasy of non-castration achieved through capitalist consumption – as the rational product of dry mathematical processing. In addition, LLMs prefigure a psychotic form of potential nonhuman subjectivity in their failure to recognise gaps in their data processing, and in their demonstration that the unconscious no longer requires a conscious human agent. The ways that LLMs relate to symbolic castration are therefore superpositioned within a simultaneously perverse and psychotic context which produces what might be called an (ordinary) perverse psychosis.
Citation
Geal, R. (2025) A large language model is structured like the unconscious: the (ordinary) perverse psychosis of AI. European Journal of Psychoanalysis, 12 (1).
Publisher
Research Unit
DOI
PubMed ID
PubMed Central ID
Embedded videos
Type
Journal article
Language
en
Description
This is an author's accepted manuscript of an article publshed in The European Journal of Psychoanalysis on 11 November 2025. The published version of the article can be accessed here: https://www.journal-psychoanalysis.eu/articles/a-large-language-model-is-structured-like-the-unconscious-the-ordinary-perverse-psychosis-of-ai/.
Series/Report no.
ISSN
2284-1059
EISSN
2284-1059