In this post Annick Backelandt argues that xenophobia should be understood as a distinct bias in Large Language Models, rather than being subsumed under racial bias. She shows how LLMs reproduce narratives of “foreignness” that particularly affect migrants and refugees, even without explicit racial references.
Shawn raises his hand and asks quietly: “Mr Warner?” […] Mr Warner does not hear Shawn or notice his raised hand. Instead, Mr Warner is fielding questions from a group of middle-class students […] Shawn sighs and puts his hand down (Calarco 2018: 164).
Post by Leonie Smith and Alfred Archer
Introduction
When middle-class students are regularly heard in the classroom and working-class students, such as Shawn, are regularly not heard, and when news reporters consistently fail to seek out women experts to the same extent that they seek out men experts, something unjust is happening. In a recent paper, we argue that this something is an epistemic attention deficit. (more…)
