People Who Know AI Won’t Necessarily Use It

Why do some people embrace AI enthusiastically while others don’t? A paper on this topic revealed a surprising and counterintuitive finding: lower AI literacy predicts greater receptivity to AI. In contrast, those with more knowledge about how AI works are less likely to embrace it.

This insight comes from a report titled “Lower Artificial Intelligence Literacy Predicts Greater AI Receptivity,” which drew on two datasets and six additional studies involving thousands of U.S.-based participants.

Enthusiasm fuels adoption

As noted by a Harvard Business Review (HBR) report, completing tasks with AI can feel magical and awe-inspiring to those who know less about how it works—this sense of wonder fuels enthusiasm. Conversely, for those with higher AI literacy who understand how algorithms, data training, and computational models function, the mystique tends to fade.

These more informed users often take a measured, less emotionally driven view of AI. According to HBR, this can result in greater caution – or even disinterest – because AI feels less novel or transformative.

This effect is especially noticeable in creative and emotional domains, where AI’s capabilities appear almost magical. People with lower AI literacy are more likely to hand over control in these areas. Interestingly, this pattern fades, or even reverses, when the task shifts to more logical domains like number crunching or data analysis, where the sense of awe is diminished.

These findings challenge a common assumption in tech adoption: that more education naturally leads to greater uptake. With AI, the opposite may be true – greater knowledge can reduce interest in AI-powered products and services.

Tailor to AI literacy level

The findings upend traditional strategies used to drive AI adoption, whether in the workplace or the marketplace. Instead of focusing on education, organizations should first assess their audience’s AI literacy using surveys, interviews, or behavioral indicators such as product usage patterns.

For AI-savvy users, the report recommends skipping the “wow factor” and focusing instead on functionality, performance, and practical outcomes. But when addressing less AI-savvy audiences, it may be better not to demystify the magic with too many technical details. But for the average, less AI-savvy audiences, it may be better not to demystify the magic with too many technical details.

In a nutshell, businesses can’t assume their most tech-savvy users are the easiest to convert. While certain use cases such as using Copilot or Cursor to write better code, are ideal for advanced users, the broader adoption of AI tools hinges on matching messaging and experience to the user’s level of AI literacy.

The paper is available here.

Image credit: Thinkhubstudio