The exponential growth of Generative Artificial Intelligence (AI) applications has shifted user expectations from mere technological novelty to functional integration into daily life. This research investigates the causal factors influencing user satisfaction, specifically comparing the impact of “Beauty” (User Interface and Design: UI/UX) against “Brains” (Response Reliability). Utilizing a Mining Software Repositories (MSR) methodology, the study analyzed a comprehensive dataset of 160 ,507 user reviews from the “Google Gemini” application on the Google Play Store. Natural Language Processing (NLP) techniques were employed to categorize key thematic clusters and perform statistical correlation analysis with user ratings. The empirical findings reveal that while UI aesthetic deficiencies moderately impact user satisfaction, system reliability issues—specifically high latency and informational inaccuracies (hallucinations) are the primary determinants of critical dissatisfaction (1 -star ratings). These results underscore that in the ecosystem of AI software, functional stability and data integrity take strategic precedence over aesthetic appeal. This study provides a foundational framework for prioritizing software development resources in the evolving landscape of Social AI.