This research examines how interactions with Large Language Models can produce negative emotional responses that impact software engineer productivity and wellbeing.

• LLM interactions in coding and requirements tasks can lead to frustration when models produce hallucinations or verbose responses
• Continued frustration may escalate to stress and burnout among engineers
• Research provides guidelines for improving human-LLM interaction in technical contexts
• Understanding these emotional patterns is crucial for effective LLM integration in engineering workflows

Emotional Strain and Frustration in LLM Interactions in Software Engineering