EUVIMEDEUVIMED entdecken
Crossref

The Writing Lesson Revisited: AI [and | in | is] the Scriptural Economy

Matthew Kirschenbaum

Social Research: An International Quarterly · 2026-06 · 93 (2) · 285-308

DOI: 10.1353/sor.2026.a992284 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Social Research: An International Quarterly, 2026-06. Verfasst von Matthew Kirschenbaum.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This essay uses Michel de Certeau's framing of modernity as a scriptural economy—and that concept's parafictional deployment in Tom McCarthy's 2015 novel Satin Island and the novel's debt to Claude Lévi-Strauss's famous chapter "A Writing Lesson"—to situate large language models (LLMs) in a tradition of writing and textual production rooted in capitalism and transactional practices. It argues that generative artificial intelligence should be understood not as the disruption of writing or an aberration but as the completion of the structural logic of the scriptural economy and that LLMs are very likely the final word on a system of production wherein writing is tokenized and transacted for extraction of value in ways that render the technology's current limitations, such as a propensity for hallucinations, largely beside the point. Lest there be any misunderstanding, this is not intended as a salutary diagnosis.

Zitieren

APA

Kirschenbaum, M. (2026). The Writing Lesson Revisited: AI [and | in | is] the Scriptural Economy. Social Research: An International Quarterly, 93(2), 285-308. https://doi.org/10.1353/sor.2026.a992284

Vancouver

Kirschenbaum M. The Writing Lesson Revisited: AI [and | in | is] the Scriptural Economy. Social Research: An International Quarterly. 2026;93(2):285-308. doi:10.1353/sor.2026.a992284

BibTeX
@article{Kirschenbaum2026,
  author = {Kirschenbaum, Matthew},
  title = {The Writing Lesson Revisited: AI [and | in | is] the Scriptural Economy},
  journal = {Social Research: An International Quarterly},
  year = {2026},
  volume = {93},
  number = {2},
  pages = {285--308},
  doi = {10.1353/sor.2026.a992284}
}
Kostenlos starten →