Visual design features can activate meanings that extend beyond their immediate perceptual properties. Building on research on visual symbolism, grounded cognition, and conceptual metaphor, we examine whether visual sharpness is associated with intelligence and how this association manifests across different levels of consumer processing. Across four studies, we provide convergent linguistic, perceptual, and associative evidence. Study 1 documents recurring sharpness–intelligence correspondences across 12 languages. Study 2 examines explicit perceptions across consumer products and category-neutral shapes and shows that the association is difficult to detect when competing product cues are available, but becomes more observable when such cues are reduced. Studies 3 and 4 use Implicit Association Tests and demonstrate greater response compatibility between sharp forms and intelligence-related concepts than between sharp forms and unintelligence-related concepts, replicating the pattern across student and general-population samples. Together, the findings suggest that visual sharpness can function as a symbolic cue that activates intelligence-related meaning, while its influence on explicit judgment depends on the informational context. The research extends visual symbolism by demonstrating how a basic geometric property can become connected with an abstract cognitive concept and distinguishes symbolic accessibility from the diagnostic role of visual cues in consumer judgment.
Consumers increasingly form what feel like social bonds with AI, yet little is known about whether or why they empathize with AI despite recognizing its artificial nature. We conceptualize human–AI empathy as consumers' tendency to cognitively understand, emotionally connect, and morally concern with AI agents. Drawing on mind perception theories and the literatures on empathy and anthropomorphism, we develop and validate the Human–AI Empathy Scale (HAES), comprising 22 items and 3 dimensions. Across three studies, we establish its reliability and convergent, discriminant, and nomological validity. We further show that tendency to anthropomorphism and mind perception are proximal antecedents of human–AI empathy, which in turn predicts satisfaction, AI attachment, willingness to pay, and forgiveness following AI service failure. HAES enables firms to segment consumers by empathic orientation toward AI, informing AI service design, targeted marketing, and strategies for cultivating enduring human–AI relationships.