1 paper · 1 filter
Bianca Helena Ximenes
This paper introduces two constructs to describe, as far as we know, a previously unnamed risk in human-LLM interaction. Co-construction blindness is the failure to recognize that…