You do not currently have permission to view the full text for this document. Please sign in to an account with an active subscription or contact support.
This essay examines the structural and conceptual differences between poetic intelligence and the linguistic abilities of artificial intelligence, in particular so-called Large Language Models (LLMs). Our starting point is the thesis that poetic intelligence is essentially determined by the capacity for negativity and the experience of absence – a capacity that LLMs structurally lack. The authors analyze Ulrich Moser's poems and compare them with the functioning of transformer-based language models, drawing on central concepts from psychoanalysis (Freud, Lacan) and metaphor theory (Blumenberg, Badiou, among others). The essay works out that in poetic language and theoretical precision LLMs can only generate positive, associative patterns, while the poem is a place of the »machinery of negativity« – a practice of thinking without an object, which is constituted by emptiness, subtraction, and metaphorical structure. Poetic intelligence is thus understood as an inhuman, radically different way of thinking that goes beyond the limits of machine language processing.
The full text is available in German.