Confidence can sound like expertise, especially in fast-moving conversations about artificial intelligence. But a Harvard survey suggests that confidence and knowledge do not always line up: some respondents said they recognized people who were invented for the study.
A test of who knows the names
Researchers asked more than 5,400 Americans whether they knew several people associated with AI. Among male respondents, 73 percent said they knew OpenAI CEO Sam Altman. A little over half, or 53 percent, said they knew Mira Murati, Susan Milner and Craig Dalton.
Murati is real: she is the former CTO of OpenAI and the current CEO of Thinking Machine Labs. Milner and Dalton, however, do not exist. The survey therefore offers a striking example of people expressing recognition even when the names include fictional figures.
Women were less likely to claim familiarity with the names, though the gap was not limited to men. Forty percent of women said they knew Milner and Dalton, and 38 percent said they recognized Murati.
Knowing more, and saying “not sure” more often
The report found that men use AI more frequently and, on average, answer factual questions about AI correctly more often. Yet men and women got questions wrong at around the same rate. One difference was how often respondents chose uncertainty: women said they were unsure on an average of 3.3 questions out of 12, compared with 2.3 for men.
That distinction matters when interpreting a survey. A higher number of correct answers can reflect greater knowledge, while fewer “not sure” responses can also signal greater willingness to guess. The results show both patterns: men performed better on average, and they were less likely to say they did not know.
Laura Manley, executive director of the Shorenstein Center at Harvard Kennedy School, told WIRED that men appear more optimistic and confident across the board when it comes to AI. She also noted that men tend to be more overconfident in surveys generally. In this case, claiming familiarity with made-up executives makes that broader tendency concrete.
Who gets heard in AI coverage?
The researchers also examined whether confidence might shape the public conversation. Across a year of media coverage, men appeared in AI headlines more than 14 times as often as women. Because senior AI executives are disproportionately men, the researchers separately looked at people quoted as subject experts.
Three-quarters of those quoted experts were men. That share was higher than the roughly 29 percent of the AI talent pool made up of women. The comparison raises a question about who is treated as a visible authority when news organizations explain AI.
Manley suggested that more women could be featured prominently as experts, including when they express uncertainty. Her point is not that uncertainty replaces expertise. Rather, making room for it may help audiences hear a fuller range of views in a field where many questions about appropriate uses and future effects remain unanswered.
Concern is shared, but optimism differs
Respondents broadly shared concern about AI: both men and women rated concern as their top emotional response. Women reported more concern, fear and worry than men, while men rated positive feelings—including curiosity, enthusiasm and excitement—higher.
Those differences sit alongside a wider divide in confidence. Public discussion can become dominated by certainty about either AI’s dangers or its promise, Manley said. When confident claims crowd out people who are willing to say they are unsure, the conversation risks making unresolved questions seem settled.
The survey points to a practical lesson for AI coverage and discussion: distinguish knowing from assuming, and make space for people to acknowledge what remains unclear. Saying “not sure” can be a useful part of understanding a technology whose uses and consequences are still being worked out.