Jianhong Luo

The authenticity penalty: Why AI-enhanced performance diminishes relational value in the creator economy

Using CLIP-based measures of text–image mismatch, it is found that the penalty appears directionally stronger when polished text is less congruent with the accompanying visuals; although the strictest interaction tests are power-limited, the marginal-effects pattern is consistent with multimodal dissonance.