Scholarlink

<p>This article examines the linguistic framing of Large Language Models (LLMs) such as Gemini and ChatGPT in the context of human-machine interaction. While human characteristics have traditionally been attributed to living beings, the discourse around LLMs reveals an unsurprising trend towards anthropomorphizing these technologies. Crucially, the analysis shows that it is not only humans who ascribe human-like traits to machines; the language models themselves actively contribute to this impression by using self-referential pronouns and verbs typically associated with human actions.</p><p>The linguistic strategies employed by LLMs blur the line between fact and mediated reality, which can potentially mislead users. This raises ethical concerns, especially regarding the perception of LLMs as conscious entities. Researchers interpret this phenomenon as a&#xa0;form of »digital animism« that affects our understanding of reality and complicates the distinction between human and artificial intelligence. The article analyses the linguistic features and strategies present in the text generated by LLMs, discusses their implications, and considers the broader impact of these design choices on the social perception of AI.</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

Das definierende Sprachmodell (LLM): Anthropomorphisierung in der Mensch-Maschine-Interaktion

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年4月11日

查看全文
  • Ekkehard Felder,
  • Marcel Kückelhaus

摘要

This article examines the linguistic framing of Large Language Models (LLMs) such as Gemini and ChatGPT in the context of human-machine interaction. While human characteristics have traditionally been attributed to living beings, the discourse around LLMs reveals an unsurprising trend towards anthropomorphizing these technologies. Crucially, the analysis shows that it is not only humans who ascribe human-like traits to machines; the language models themselves actively contribute to this impression by using self-referential pronouns and verbs typically associated with human actions.

The linguistic strategies employed by LLMs blur the line between fact and mediated reality, which can potentially mislead users. This raises ethical concerns, especially regarding the perception of LLMs as conscious entities. Researchers interpret this phenomenon as a form of »digital animism« that affects our understanding of reality and complicates the distinction between human and artificial intelligence. The article analyses the linguistic features and strategies present in the text generated by LLMs, discusses their implications, and considers the broader impact of these design choices on the social perception of AI.