Who is asking the machine

The education gap is enormous and it is measured, not guessed. Among U.S. adults with a postgraduate degree "52% say they have used ChatGPT," against "18% with a high school degree or less education" (Pew Research Center, June 25, 2025). Nearly three to one.

Income follows the same slope, with one honest caveat. Economists Bick, Blandin and Deming, in a Federal Reserve Bank of St. Louis working paper, found that "Adoption has been faster for male, younger, more educated, and higher-wage workers," that "workers with a Bachelor's degree or more use genAI at roughly twice the rate of non-college educated workers," and that adoption rates "increase steadily with income until roughly the 80th percentile" (Bick, Blandin & Deming, St. Louis Fed Working Paper 2024-027, revised October 2025). Read that last clause carefully, because we would rather you heard the limit from us than from a competitor: the curve flattens at the top. The heaviest users are the well-paid professional class, not the very top of the distribution, and that professional class is exactly who signs off on a vendor.

Usage clusters geographically, and the clusters are not where the headlines would put them. Anthropic's September 2025 index of Claude usage per working-age person found that inside the United States "the District of Columbia leads with an Anthropic AI Usage Index of 3.82" followed by "Utah (3.78), notably ahead of California (2.13)." Globally, "Israel leads global per capita Claude usage with an Anthropic AI Usage Index of 7" and "Singapore follows at 4.57," with a "strong positive correlation between Claude adoption and Gross Domestic Product per working-age capita" (Anthropic Economic Index, September 15, 2025).

The early skew is also correcting, fast. The NBER study of ChatGPT's first two and a half years found that "Early adopters were disproportionately male but the gender gap has narrowed dramatically, and we find higher growth rates in lower-income countries" (Chatterji et al., NBER Working Paper 34255, September 2025). Whatever the audience looks like today, it looks more like everyone tomorrow.

Put the pieces together and the picture is stark. The people most likely to ask an answer engine which product to buy are under fifty, college educated, earning above the median, concentrated in the richest metros, and asking daily. That is not a demographic you can afford to be invisible to. It is the demographic every pitch deck claims to be targeting, and the machine has become its research assistant.

We measure exactly that: the questions those buyers ask, on the engines they use, with live search on, sampled and stored so you can read the answers yourself. Half of America is already asking. Let's find out what the machine says when they ask about you: the scorecard is free.