Scholarlink

<p>In this paper I attempt to critique some recent accounts of how AI systems can (not) inhabit our space of reasons, using Robert Brandom’s inferentialist semantics. In particular, I argue that the extreme complexity of deep neural networks does not prevent AI systems based on them from qualifying as rational agents, and that a given AI system can inhabit the space of reasons if it can play the game of giving and asking for reasons. In the last part of my paper, I argue that rethinking the place of AI systems in the space of reasons leads to a reopening of the debate about the transparency/explainability of algorithms.</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

There Are No Aliens in the Space of Reasons

  • 期刊论文

  • 发表日期：

    2025年7月9日

查看全文
  • Leonardo Marchettoni

摘要

In this paper I attempt to critique some recent accounts of how AI systems can (not) inhabit our space of reasons, using Robert Brandom’s inferentialist semantics. In particular, I argue that the extreme complexity of deep neural networks does not prevent AI systems based on them from qualifying as rational agents, and that a given AI system can inhabit the space of reasons if it can play the game of giving and asking for reasons. In the last part of my paper, I argue that rethinking the place of AI systems in the space of reasons leads to a reopening of the debate about the transparency/explainability of algorithms.