AI Literacy Predicts Non-Text Tool Adoption, Not General Receptivity

Previous research suggested a general inverse relationship where lower artificial intelligence (AI) literacy correlated with greater receptivity towards AI. This paper re-examines that claim by re-analyzing public data from a study that measured past usage across five distinct categories of AI tools. The re-analysis confirms the negative association between AI literacy and aggregate AI usage. However, it uncovers significant heterogeneity when breaking down usage by tool type. The study found that AI literacy did not significantly predict the usage of text-based AI tools. Instead, it remained a strong predictor for the broader adoption of non-text AI tools. Furthermore, the observed relationship primarily indicates an adoption/non-adoption pattern, meaning lower AI literacy is linked to having 'ever used' a non-text AI tool, rather than demonstrating intensive or frequent use. This suggests that the connection between AI literacy and AI engagement is more nuanced and tool-specific than a simple, generalized claim of higher receptivity among those with lower literacy.