CONCEPT
Inferential Literacy
The disciplined ability to separate the quality of an AI’s presentation from the accuracy of its content—the cognitive reflex that asks, of every output, whether the confidence of the prose matches the warrant of the evidence.
Edward Tufte spent a career teaching visual literacy: the trained ability to read a graphic with the same critical attention one brings to reading an argument, to ask of every chart whether the effect shown in the display matches the effect in the data. The AI transition has made an analogous discipline urgently necessary—one that applies to language rather than to graphics, and that Tufte himself demonstrated in his July 2025 response to a widely shared claim about Microsoft’s AI diagnostic framework. Inferential literacy is the reflex to ask, of every AI-generated output, what the
lie factor of its prose is: the ratio between the confidence of the presentation and the accuracy of the content.
Large language models are engagement-optimized systems trained on vast corpora of human text, and the patterns they have absorbed include not only the structures of accurate communication but also the structures of persuasive, confident, polished communication—all of the rhetorical modes humans deploy when the priority