Decoding algorithm fatigue: The role of algorithmic literacy, information cocoons, and algorithmic opacity

Abstract

Algorithmic technologies are dominating our online experiences, from content recommendations to personalized services. However, it also introduces a new challenge: algorithm fatigue. Algorithm fatigue describes the phenomenon where users experience mental and emotional exhaustion in prolonged interaction with algorithms. To decode this phenomenon, we explored personal and technical antecedents of algorithm fatigue and its impact on user behavior. Using data collected from 393 users of algorithm-driven applications, we identified three key drivers of algorithm fatigue: algorithmic literacy (understanding how algorithms work), information cocoons (being exposed to repetitive content), and algorithmic opacity (a lack of transparency in how algorithms make decisions). Interestingly, while greater algorithmic literacy is often thought to enhance user satisfaction, our findings suggest it actually exacerbates fatigue. Both information cocoons and algorithmic opacity also contribute to algorithm fatigue, highlighting the need for diverse content and transparent algorithm designs. Additionally, we found a strong link between algorithm fatigue and resistance behavior, with fatigued users more likely to resist algorithmic recommendations. Overall, this study suggests developers and policymakers design more user-centric algorithms that not only excel in personalization but also reduce potential fatigue and resistance in algorithmic interactions.