Powered by OpenAIRE graph
Found an issue? Give us feedback
addClaim

ChatGPT

Banale Täuschung und spektrale Autorschaft
Authors: Hogrebe, Vincent;
Abstract

Die Arbeit widmet sich den Implikationen von Large Language Models (LLMs) wie ChatGPT auf Konzepte von Autorschaft. Der zentrale analytische Rahmen bildet das Konzept der banalen Täuschung von Simone Natale, welches beschreibt, wie illusionäre Funktionalität und Anthropomorphismen die Wahrnehmung von KI als intelligent ermöglichen. Es wird dargelegt, dass LLMs Sprache auf Basis einer strukturalistischen Logik der langue nutzen, welche die Fähigkeit zur sprachlichen Kohärenz aus einem Differenzsystem zieht. Im Gegensatz zur menschlichen Kommunikation fehlt der maschinell generierten Poesie der lebensweltliche Bezug und die Intentionalität der parole, was in Bezug auf ein monolithisches Autorschaftsmodell zu einer repräsentativen Leerstelle führt. Die Arbeit beleuchtet einige Facetten dieser Repräsentationskrise. Einerseits wird die Sinnkonstitution maßgeblich der lesenden Person zugeschrieben, die im performativen Akt der Lektüre Bewusstseinskorrelate synthetisiert und dadurch ästhetische Bedeutung generiert. Andererseits trägt das Interface als black box zur Täuschung bei, indem es technische Arbeitsprozesse verschleiert. Durch die echtzeitliche Interaktion erzeugt das Interface die Illusion einer „appresent presence“, in der ein technisches Anderes sozial interagiert, sich jedoch nie physisch materialisiert. Zur Charakterisierung der LLM-Autorschaft führe ich abschließend das Konzept der spektralen Autorschaft ein. Dieses Konzept integriert die identifizierten Täuschungsmechanismen und fasst die Kreation künstlicher Poesie in einem Netzwerk technischer, medialer, psychosozialer und kunsthistorischer Bedingungen. Spektrale Autorschaft beruht auf einer Irritation des chronologischen Vektors, indem sie einen aktualisierenden Rückbezug der Vergangenheit durch Textspuren („Nicht-mehr“) mit der antizipierten, illusorischen Präsenz des Anderen („Noch-nicht“) im Augenblick der virtuellen Interaktion verschränkt.

This paper examines the implications of large language models (LLMs) such as ChatGPT on concepts of authorship. The central analytical framework is Simone Natale's concept of banal deception, which describes how illusory functionality and anthropomorphism enable the perception of AI as intelligent. It is argued that LLMs use language based on a structuralist logic of langue, which derives its capacity for linguistic coherence from a system of differences. In contrast to human communication, machine-generated poetry lacks the reference of a lived experience and intentionality of parole, which leads to a representative void in relation to a monolithic model of authorship. The work highlights some aspects of this crisis of representation. On the one hand, the constitution of meaning is largely attributed to the reader, who synthesizes correlates of consciousness in the performative act of reading and thereby generates aesthetic meaning. On the other hand, the interface poses a black box by obscuring technical work processes. Through real-time interaction, the interface creates the illusion of an “appresent presence” in which a technical other interacts socially but never physically materializes. To characterize LLM authorship, I conclude by introducing the concept of spectral authorship. This concept integrates the identified mechanisms of deception and summarizes the creation of artificial poetry in a network of technical, media, psychosocial, and art-historical conditions. Spectral authorship is based on an irritation of the chronological vector by intertwining an actualizing reference back to the past through textual traces (“no longer”) with the anticipated, illusory presence of the other (“not yet”) at the moment of virtual interaction.

Keywords

Machine Learning, 028 Lesen und Nutzung anderer Informationsmedien, Large Language Models, Hauntology, Artificial Intelligence, Künstliche Intelligenz, Digital Literature, Digitale Literatur, Hantologie, Maschinelles Lernen

  • BIP!
    Impact byBIP!
    selected citations
    These citations are derived from selected sources.
    This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    0
    popularity
    This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
    Average
    influence
    This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    Average
    impulse
    This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
    Average
Powered by OpenAIRE graph
Found an issue? Give us feedback
selected citations
These citations are derived from selected sources.
This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Citations provided by BIP!
popularity
This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
BIP!Popularity provided by BIP!
influence
This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Influence provided by BIP!
impulse
This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
BIP!Impulse provided by BIP!
0
Average
Average
Average
Related to Research communities
Upload OA version
Are you the author of this publication? Upload your Open Access version to Zenodo!
It’s fast and easy, just two clicks!