Scholarlink

<p>This articles replies to Tiegue Vieira Rodrigues’s commentary “The Epistemology of Algorithmic Narrative and the Problem of Creative Authenticity” on Mollema’s “AI-generated Literature, Distant Writing and the Reader: Reflections on Floridi and Calvino”. I appreciate the opportunity of engaging in discussion with Rodrigues, a fellow ‘early reflector’ on Luciano Floridi’s concept of ‘distant writing’. Rodrigues and I agree on my core claims (a) that the reader’s interpretative role with respect to AI-generated literature remains constitutive for AI-generated literature’s meaning; and (b) that AI’s narrative capabilities are not exceptional, but stand in continuity with the human author’s literary production. But Rodrigues argues against (b) that this de-anthropomorphized conception of literary production glosses over a major epistemological difference: Large Language Models (LLMs) do not understand their navigation through narrative space, while human authors do. It is this concern that I shall address here. I object to Rodrigues’s conclusion that we can still account for psychological differences in creation (i.e., human authors interpret themselves, whereas LLMs do not), while maintaining a structural equivalence on a specific functional domain. Furthermore, I express Wittgensteinian doubts regarding the dependence of his argument on the concept of understanding.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

On Large Language Models, De-anthropomorphized Narrative Production, and Deflationary Understanding: A Reply to Rodrigues

  • 期刊论文

  • 发表日期：

    2026年1月3日

查看全文
  • Warmhold Jan Thomas Mollema

摘要

This articles replies to Tiegue Vieira Rodrigues’s commentary “The Epistemology of Algorithmic Narrative and the Problem of Creative Authenticity” on Mollema’s “AI-generated Literature, Distant Writing and the Reader: Reflections on Floridi and Calvino”. I appreciate the opportunity of engaging in discussion with Rodrigues, a fellow ‘early reflector’ on Luciano Floridi’s concept of ‘distant writing’. Rodrigues and I agree on my core claims (a) that the reader’s interpretative role with respect to AI-generated literature remains constitutive for AI-generated literature’s meaning; and (b) that AI’s narrative capabilities are not exceptional, but stand in continuity with the human author’s literary production. But Rodrigues argues against (b) that this de-anthropomorphized conception of literary production glosses over a major epistemological difference: Large Language Models (LLMs) do not understand their navigation through narrative space, while human authors do. It is this concern that I shall address here. I object to Rodrigues’s conclusion that we can still account for psychological differences in creation (i.e., human authors interpret themselves, whereas LLMs do not), while maintaining a structural equivalence on a specific functional domain. Furthermore, I express Wittgensteinian doubts regarding the dependence of his argument on the concept of understanding.