Designing for the Invisible: Raising Awareness for the Experience of the Homeless in an LLM-based Serious Game

The experiences of the homeless are often distorted in public discourse, leading to stereotypes that overshadow underlying motives. To give voice to the perspectives of the homeless, we designed an interactive serious game based on in-depth interviews with eight homeless individuals. Qualitative coding of the transcripts revealed key themes that guided the design of three NPCs, whose dialogues were represented using LLMs. User testing revealed that players actively initiated conversations with LLM-based NPCs, used empathetic language during dialogues, and demonstrated empathy during the dialogues. However, LLMs struggle to capture the complexity of homeless people fully, sometimes misrepresenting sensitive experiences, reducing authenticity of the experience. These findings suggest first-person interview strategies for careful consideration of limitations of LLMs in attempting to represent marginalized groups. We deliberate on the scope of serious games for raising awareness of social issues while stating their biases and limitations in representing marginalized groups.

Institutions

Publication Details

Published
2025-09-03
DOI
https://doi.org/10.1145/3748699.3749822
Citations
1
Primary Topic
Homelessness and Social Issues
Type
article
Field-Weighted Citation Impact
2.82
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Designing for the Invisible: Raising Awareness for the Experience of the Homeless in an LLM-based Serious Game

1 citations
Homelessness and Social Issues
2.82
article

Designing for the Invisible: Raising Awareness for the Experience of the Homeless in an LLM-based Serious Game

article en
1 citations

Abstract

The experiences of the homeless are often distorted in public discourse, leading to stereotypes that overshadow underlying motives. To give voice to the perspectives of the homeless, we designed an interactive serious game based on in-depth interviews with eight homeless individuals. Qualitative coding of the transcripts revealed key themes that guided the design of three NPCs, whose dialogues were represented using LLMs. User testing revealed that players actively initiated conversations with LLM-based NPCs, used empathetic language during dialogues, and demonstrated empathy during the dialogues. However, LLMs struggle to capture the complexity of homeless people fully, sometimes misrepresenting sensitive experiences, reducing authenticity of the experience. These findings suggest first-person interview strategies for careful consideration of limitations of LLMs in attempting to represent marginalized groups. We deliberate on the scope of serious games for raising awareness of social issues while stating their biases and limitations in representing marginalized groups.

University of Auckland (NZ), City University of Hong Kong (HK), New York University (US), Carnegie Mellon University (US)
Reduced inequalities
Openalex Percentile: Top 7%
Homelessness and Social Issues
2.82
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.