Human-AI Ranking Aggregation

Jonas Karge
Human-AI Ranking Aggregation
Talk, 2026

KurzfassungAbstract
In this talk, we present ideas for a collaborative project on human–AI ranking aggregation. More specifically, we explore how panels of human agents and AI systems can combine their judgments in a reliable way when their belief states are represented as rankings over possible outcomes. We begin by illustrating the aggregation mechanism through simple pooling rules that yield a collective ranking. We then examine common human–AI interaction patterns and the pitfalls they may involve. Our central goal is to identify when a given interaction pattern and aggregation rule enable the group to make better decisions than either the humans or the AI would make on their own. Finally, we outline the key questions that need to be answered in order to achieve this goal.
Projekt:Project: SECAI
Forschungsgruppe:Research Group: Computational LogicComputational Logic

@misc{K2026,
  author = {Jonas Karge},
  title  = {Human-AI Ranking Aggregation},
  year   = {2026}
}

This presentation was given at the AIRU-KRR seminar held at the University of Cape Town on March 25, 2026.